Introduction

Viral reactivation is associated with different types of stimuli including physiologic and physical changes, and in particular, immunosuppression [1,2,3]. Viral reactivation consists of two cycles, including lytic and latent phases. Within lytic cycle, replication and expression of a viral genome and release of virions result in the lysis of host cells whereas during latency, some human viruses may remain dormant within host cells and establish persistent infection with limited or no production of viral particles [1,2,3]. The pathogens may switch between the latent and lytic cycles, and a process in which a latent virus enters the lytic stage is known as reactivation [1,2,3].

Patients with severe COVID-19 have been characterized by impaired immunity, hyperinflammation, lymphopenia, and cytokine storms [4]. Decline in a number of CD4+ and CD8+ T cells due to either direct attack from SARS-CoV-2 through the spike (S) protein binding with a receptor on T cell [4,5,6,7] or induced cellular apoptosis and the subsequent decline in the type I interferons, especially interferon-gamma (IFN-γ) [4, 8, 9], as well as the T cell exhaustion marked by the increased expression of programmed death 1 (PD-1) indicate the state of immunosuppression in COVID-19 patients [4, 10]. Moreover, elevated levels of IL-2 and TNF-α can mediate T cell apoptosis by promoting Fas signaling and exacerbate lymphopenia in COVID-19 patients [4, 11,12,13]. An impaired host immune system without a normal suppression of virus replication, consequently, may induce the reactivation of latent viruses in host, which could contribute to some of the neurologic, dermatologic, and hematologic manifestations among others in COVID-19 patients [14,15,16,17,18,19,20,21,22,23,24].

Adverse clinical outcomes from the interplays between SARS-CoV-2 and other respiratory and systemic viruses have also become evident from their synergistic impact on further increase of the inflammatory cells to the site of infection and elevation of proinflammatory cytokines [25,26,27]. For example, patients with SARS-CoV-2 and influenza virus coinfections were found to develop hyperinflammation, ARDS, myocarditis, acute kidney injury, and other disorders due to more frequent activation of the cytokine cascade by flu infection [28,29,30,31,32]. A retrospective study by Yue et al. revealed that compared to patients with COVID-19 alone, those coinfected with SARS-CoV-2 and Influenza B virus (IBV) were more likely to have poorer prognosis marked by the fatigue, abnormalities in chest computed tomography (CT), and decreased lymphocytes and eosinophils, however these findings were not consistent in patients with Influenza A Virus (IAV) [33]. Primary infection by a different virus, such as cytomegalovirus, followed by SARS-CoV-2 infection, may also predispose individuals to more severe COVID-19 by compromising immunity through disruption of T-cell differentiation and upregulation of interleukin-6 [28]. immunocompromised individuals could be more prone to viral coinfections as Lino et al. observed a higher prevalence of therapeutic immunosuppression status among HHV-6/SARS-CoV-2 coinfected patients compared to the SARS-CoV-2 patients only [34]. However, exact mechanisms by which secondary infections with viruses occur in COVID-19 patients are under investigation.

Studies have additionally reported the incidence of reactivated viruses in COVID-19 patients, mainly the Herpesviridae virus including herpes simplex type 1 and 2 (HSV-1 and HSV-2), varicella zoster virus (VZV), Epstein–Barr virus (EBV), cytomegalovirus (CMV), human herpes virus 6, 7 (HHV 6, 7) [14,15,16,17,18,19,20,21,22,23,24]. In rare cases, hepatitis B virus (HBV) reactivation in chronic patients have also been observed [35, 36]. le Bal’ch et al. reviewed the virology results of 38 COVID-19 patients and found the patients with either CMV or HSV reactivation required prolonged mechanical ventilation compared to patients with no reactivated Herpesviridae virus [15]. Similarly, Simonnet et al. performed systematic testing for EBV, CMV, HHV-6 DNAemia on critically ill COVID-19 patients and observed the virus reactivations in 85% of patients, among which the patients with EBV reactivation required longer hospital length-of-stay [17]. Skin and ocular manifestations, too, including lesions in multiple regions and herpetic keratitis with risk of blindness, have been observed in patients with Herpesviridae reactivations [18, 19]. Liu et al. reported HBV reactivation in three COVID-19 patients with a history of chronic HBV infection where an increase in HBV DNA upon admission was noted, with one patient developing cirrhosis [35].

Albeit the emerging evidence present a broad spectrum of medical complications among the patients, there is lack of knowledge on latent virus reactivation in COVID-19 cases. Published studies narrowly focus on specific virus species rather than conducting an overarching investigation on all relevant latent virus activation and coinfection in patients. To increase the effectiveness of complex patient care and prevent potential deterioration of patient condition from infections during COVID-19, it is critical to establish a solid understanding of the consequences of both virus reactivation and coinfections. Thus, through the systematic review, we attempted to investigate what were the reported clinical and demographic characteristics of patients undergoing either latent virus reactivation or coinfection, what diagnostic tools were used to detect the reactivation and coinfection, and whether any COVID-19 treatment was associated with an occurrence of either reactivation or coinfection, or both.

Methods

This study was reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement (see Additional file 1 for the checklist) [37]. A predetermined study protocol of data sources, search strategies, inclusion criteria, and data extraction method was registered on the International Prospective Register of Systematic Reviews (PROSPERO, registration number: CRD42022340897) available at https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022340897.

Population of interest

The primary population of interest consisted of all patients with confirmed COVID-19 diagnosis via reverse transcription polymerase chain reaction (RT-PCR), medical and admission history, or laboratory confirmed positivity if not specified otherwise, who were diagnosed with a viral infection either concurrently or following the COVID-19 diagnosis. There were no age restrictions.

Data sources

We extracted the relevant literature through a systematic search using the key terms such as viruses, COVID-19, reactivation, latent infection, resurrection, coinfection in the online databases including the EMBASE, MEDLINE, Latin American Caribbean Health Sciences Literature (LILACS), from December 2019 up to June 2022. Manual search was also performed using the Google Scholar for identifying additional articles. Specific key terms for each database are listed in Appendix S1 in Additional file 2.

Study selection

The review included cohort studies, cross-sectional studies, case reports, case series, preprints, and editorial letters with relevant cases described. Studies published in English, Portuguese, Spanish, French, and Korean were sought for. There were no restrictions on the country and age of the population. Studies targeting irrelevant population or animals, describing a case with suspected but no confirmed diagnosis without molecular testing, as well as the literature in the formats of commentary, survey, and recommendation were excluded from the review. Two investigators (YHK, LR) separately screened the titles and abstracts of articles and simultaneously resolved the conflicts (Fig. 1).

Fig. 1
figure 1

PRISMA flowchart

Full size image

Data extraction

We used MS Excel as a template for data extraction. Following information were extracted from each article: first author, publication year, study title, study site, study setting, study design, secondary infection type (reactivation, coinfection, not distinguished), total number of participants, total number of COVID-19 positive participants, total number of COVID-19 positive participants with secondary infection, age, sex/sex ratio, medical history and existing comorbidities, laboratory findings, clinical manifestations including symptoms and complications, secondary infection definition/criteria described in a study, diagnostic tools, treatments and procedures received.

Risk of bias assessment

YHK, MT, FT, and MR independently assessed the risk of bias for cohort studies. YHK independently assessed the risk of bias for case reports. Meetings were arranged between authors to overlook and resolve any disagreements on the assessments. For the case reports and case series, the Consensus-based Clinical Case Reporting (CARE) guidelines will be used. For the cohort studies, the Newcastle–Ottawa Scale (NOS) will be used.

Summarizing data

For cohort and cross-sectional studies that do not provide distinct clinical data by each virus, only the frequency of cases was imputed in tables. Laboratory findings summarized either in mean or median values were distinguished as categorical variables either as elevated or decreased levels using the external references [38,39,40] when the reference ranges were not provided. For each relevant laboratory variable, total reactivation cohort count was imputed. For case reports describing co-reactivations in a single patient, we added a count of 1 for variables for each virus type.

Results

Study selection

From the search, total 1316 records were extracted from searches. Upon removal of duplicates (n = 160), 1156 articles were eligible for title and abstract screening, among which 1046 were excluded due to irrelevant study settings, patient populations, and not having access to full texts. 110 records were eligible for the full text screening, among which 57 were excluded for insufficient data for the objectives of this review. Thus, 53 records reporting an infection secondary to COVID-19 diagnosis were included in this study (Fig. 1). All selected articles were published in either English or available with English-translated versions. No separate translation for other languages was required in data extraction.

Study characteristics

Among 53 included studies, we found 22 case reports, 20 cohort studies, 7 case series, 2 cross-sectional studies, and 2 editorial letters in formats of 1 case report and 1 cohort study. Out of 45 studies based on inpatient settings, 24 studies reported the cases of infection in patients admitted to the Intensive Care Unit (ICU). There was 1 autopsy report, 1 Electronic Medical Record (EMR) based report, and 1 study with no specified study setting. The rest were based on ambulatory care settings (Table 1—located below the List of Abbreviations). Table 1 summarizes the characteristics of studies included in this review, including author and publication year, study site, study design and setting, total study population, age of study population expressed either in median or mean years, outcome of the study, COVID-19 diagnostic method, and diagnostic method for the outcome. Here, outcome refers to either viral co-infection or reactivation cases for which the authors provided the clinical outcomes of patients. Information on undetected viruses and patients with no additional information besides the infection status were excluded due to the nature of aims for this study.

Table 1 Study characteristics of selected records
Full size table

Risk of bias assessment

All cohort studies were categorized as either high quality (n = 13) or moderate quality (n = 8). Stars were deducted from the studies mainly from the Selection domain where the absence of outcome interest, secondary infection, at the start of study was not indicated or the non-COVID-19 cohort was not selected. 11 case reports and 3 series satisfied 70% or more of the CARE checklist items. Majority of the reports were deducted points from the Title domain without a title describing a primary diagnosis or intervention followed by the word “case report”, Abstract and Introduction domains without discussion on scientific benefits of the study, Timeline domain without an organized timeline of patient care, and Patient Perspective domain without discussion on patient’s experience with treatment(s) they received. Some case reports also did not indicate whether the patient consents were obtained. Assessment results are available on Appendix S2 in Additional file 2.

Definition of reactivation and co-infection

The review identified 40 reactivation studies, 8 studies co-infection studies, and 5 studies where concomitant infection in COVID-19 patients was not distinguished as either reactivation or co-infection. All reported cases have been categorized accordingly to authors’ descriptions in relevant studies and Table 2 summarizes the detection criteria of each virus by type of infection.

Table 2 Summary of detection criteria for each infection category
Full size table

Reactivation

Table 3 summarizes the findings for the reactivation cohorts. In total, there were 895 cases of reactivations, among which 519 (58%) were Epstein Barr virus (EBV) reactivation, 157 (17.5%) were human herpes virus 1 (HHV-1) reactivation, 154 (17.2%) were cytomegalovirus (CMV) reactivation, 47 (5.3%) were HHV-6 reactivation, 14 (1.6%) were hepatitis B virus (HBV) reactivation, 2 (0.2%) were HHV-8, and 1 (0.11%) of varicella zoster virus (VZV), HHV-2, HHV-7, and HHV-8 were observed reactivations were observed at rates less than 1% each. 7 studies reported 336 patients who underwent co-reactivations of different viruses.

Table 3 Summary of clinical findings from reactivations
Full size table

Among the cohorts whose COVID-19 severity levels were reported, 103 EBV patients, 116 CMV patients, and 150 HHV-1 patients, and 7 HHV-6 patients populations experienced severe or critical illness. Overall, hypertension (n = 70, 32.0%), immunosuppression (n = 38, 17.4%), diabetes (n = 53, 24.2%), and cardiovascular disease (n = 22, 10.1%) were most frequently observed across reactivation cohorts. In particular, patients with CMV and HHV-1 reactivations had high rates of comorbidities, where hypertension (n = 61, 39.6%), diabetes (n = 37, 2%), immunosuppression (n = 21, 13.6%), and cardiovascular disease (n = 8, 5.2%) were common among CMV patients, and similarly, immunosuppression (n = 17, 12%), diabetes (n = 13, 9.2%), and cardiovascular disease (n = 12, 8.5%) were common among HHV-1 patients. Cough and fever had high frequencies, with fever reported in all populations except in the VZV, HHV-2, and HHV-8 groups. All patient except for the HHV-7 cohort experienced complications. 108 EBV patients and 9 CMV patients reported to have ARDS, taking more than 20% of the complication frequencies.

All patient except for the HHV-7 cohort experienced complications, and death occurred in patients with EBV (n = 12, 2.3%), CMV (n = 16, 6.3%), HHV-1 (n = 31, 19.7%), and HBV (n = 1, 8.3%) reactivations. Among laboratory findings, lymphopenia was observed in all cohorts (n = 144, 13.42%) except for VZV, HHV-2, and HHV-8, and was most prevalent among CMV and HHV-1 patients. Elevated CRP level was the manifestation with the highest overall frequency (n = 207, 19.29%) from CMV (n = 89) and HHV-1 (n = 117) cohorts, next to elevated D-dimer level (n = 205, 19.11%) from EBV (n = 55), CMV (n = 88), HHV-1 (n = 62) cohorts.

When calculating out of total frequencies of pharmaceutical interventions in each virus group, acyclovir (n = 72, 17.1%) and ganciclovir (n = 43, 10.2%) were the most common forms of treatment for CMV reactivation whereas acyclovir alone was provided most frequently for controlling the HSV-1 viremia (n = 111, 57.8%). Overall, steroids (n = 176, 27.7%), including betamethasone (n = 2), dexamethasone (n = 42), methylprednisolone (n = 4), and non-specified steroids (n = 128), as well as tocilizumab (n = 87, 13.9%) were provided as part of the COVID-19 treatment across different virus cohorts.

Coinfection

Table 4 summarizes the findings for the coinfection groups. In total, 265 coinfections were reported, among which 187 (70.6%) were IAV, 23 (8.7%) were IBV, 38 (14.3%) were EBV, 13 (4.9%) were HHV-6, 3 (1.1%) were CMV, and 1 (0.04%) was VZV infected patient. None of the cohorts but IAV and CMV reported COVID-19 severity where 10 and 2 patients experienced severe or critical illness, respectively. All but IBV and VZV patient reported comorbidities and past medical history, where cardiovascular disease was observed in both EBV (n = 1) and HHV-6 (n = 8) cohorts whereas diabetes and malignancy were observed in the CMV (n = 2, n = 1, respectively) and HHV-6 (n = 8, n = 4, respectively) cohort.

Table 4 Summary of clinical findings from coinfections
Full size table

Out of total, cardiovascular disease and diabetes reported the highest frequencies with 23.68% and 26.32%, respectively. Acute kidney injury and sepsis occurred as complications among the CMV (n = 1, n = 2, respectively) and HHV-6 patients (n = 9, n = 4, respectively). Fever was reported in IAV (n = 134), IBV (n = 16), EBV (n = 27), CMV (n = 2), and VZV (n = 1) cohorts, cough in IAV (n = 39), IBV (n = 6), EBV (n = 28) and CMV (n = 2) cohorts, dyspnea in IAV (n = 13), IBV (n = 2), EBV (n = 1) and CMV (n = 2) cohorts, and myalgia in EBV (n = 18) and VZV (n = 1) cohorts. Among laboratory findings, lymphopenia was dominant in IAV cohort (n = 187), marking the highest frequency with 92.8%. Elevated D-dimer and elevated procalcitonin levels were reported in both EBV (n = 1, n = 1) and CMV (n = 3, n = 1) populations summing to the frequencies of 1.69% and 0.85% each.

For pharmacological interventions, EBV and CMV cohorts had overlaps in receiving remdesivir (n = 1, n = 1, respectively), non-specified antibiotics (n = 32, n = 1, respectively), and steroids (n = 22, n = 2, respectively). Overall, antivirals, including acyclovir, ganciclovir, lopinavir/ritonavir, oseltamivir, remdesivir, and non-specified antivirals, were the most frequently prescribed medications across the cohorts summing up to 53.68% of total interventions, among which oseltamivir was reported with the highest frequency (23.53%). Oxygen therapy and invasive ventilation were common procedures with 69.23% and 23.08% of frequencies.

Indistinguishable infection

Table 5 summarizes the findings for indistinguishable groups where the terms “reactivation” and “coinfection” were interchangeably used in articles to describe the patients. In total, 51 cases were indistinguishable. EBV cases (n = 46) were most frequently reported next to CMV (n = 3), VZV (n = 1), and B19 (n = 1). Information on COVID-19 severity was not available on the observed cohorts but EBV, where 7 patients experienced severe or critical illness.

Table 5 Summary of clinical findings from indistinguishable infections
Full size table

There was no comorbidity or medical history among patient groups except for the history of previous VZV infection in VZV and B19 groups and pneumonia in CMV and B19 groups. 1 EBV and 1 B19 patients reported rash, and 1 CMV and 1 VZV patients underwent rectal ulcer and neuralgia respectively as complications upon infections.

Both EBV and B19 cohorts experienced elevated ALT (n = 1, n = 1), D-dimer (n = 1, n = 1), and LDH levels (n = 1, n = 1). Available treatment information indicates provision of acyclovir and pregabalin to 1 VZV patient each, and valganciclovir for 1 CMV patient for controlling the viral load. Supportive oxygen therapy was most frequent, taking up 88.99% of total procedures.

Discussion

Overall, our review identified 53 full text articles from the literature search, of which 40 studies reported viral reactivations, 8 studies reported viral coinfections during COVID-19 infection, and 5 studies reported viral infections which were not distinguished as either reactivation or coinfection. Case reports and cohort studies consisted over 50% of the included studies. Among the patients reported, those admitted to the inpatient units including ICU were most common next to the patients receiving the ambulatory care. Risk of bias assessments revealed 13 high and 8 moderate quality cohort studies, 2 high quality cross-sectional studies, and 11 case reports and 3 case series satisfying 70% or more of the CARE checklist items. We did not include any study with prevalence rates available only as one of the primary objectives of this study was to present summary of clinical characteristics of patients. In total, we were able to extract data for 12 viruses including the IAV, IBV, EBV, CMV, VZV, HHV-1, HHV-2, HHV-6, HHV-7, HHV-8, HBV, and Parvovirus B19. Upon summarizing the clinical findings for each infection category, EBV, HHV-1, and CMV were most frequently observed within the reactivation cohort, IAV and EBV within the coinfection cohort, and EBV within the indistinguishable infection cohort. In both reactivation and coinfection groups, patients were commonly reported to have cardiovascular disease, diabetes, and immunosuppression as comorbidities, cough, fever, and myalgia as symptoms, acute kidney injury as complication, and lymphopenia and elevated D-dimer and CRP levels from blood tests. Common pharmaceutical interventions in both groups were steroids including dexamethasone and non-specified steroids, as well as antivirals including acyclovir, ganciclovir, remdesivir, and otherwise non-specified antivirals.

We provide two possible hypotheses for the latent virus reactivation and coinfection mechanisms. First, viral reactivation and coinfection may be attributed to impaired cellular immunity by the existing SARS-CoV-2 infection [75, 78, 84,85,86,87]. From our review, we observed high frequencies of lymphopenia with 13.42% and 92.8% in reactivation and coinfection cohorts, respectively. Similarly, retrospective study in China laboratory-confirmed 24 additional respiratory pathogens among COVID-19 patients, and 94.2% of the patients were co-infected with one or more pathogens [84]. Additionally, Paolucci et al. observed a significant loss of lymphocytes in COVID-19 patients and reported an association with the reduction of natural killer (NK) cell and CD8+ T cells and presence of EBV DNA [75]. Exhaustion of functional T cells have been observed among COVID-19 patients marked by the increased expressions of the programmed cell death protein 1 (PD-1) and T cell immunoglobulin and mucin domain 3 (Tim-3), which are indicative of the loss of effector T cell function in viral clearance [78, 85, 86]. Additionally, Ouyang et al. in their 2020 reported the downregulation of the key proteins in TCR signaling and T-cell activation, including SOS1 and MAP2K7, during SARS-CoV-2 infection [87]. Although there are accumulating evidence of diminished immunity among COVID-19 patients, specific immunopathological mechanisms of COVID-19 and viral reactivations and coinfections are yet to be established.

Second, majority of patients who were critically ill from the SARS-CoV-2 infection upon admission received immunomodulatory agents. Steroidal inhibition of inflammatory proteins through gene suppression and blockade of T cell differentiation and monoclonal antibody blockade of IL-6 receptor have been proposed to be beneficial for controlling the cytokine storm and associated organ damage [88,89,90], yet such immunosuppressant properties of treatment may had favored opportunistic infections or latent virus replications [91, 92]. This is supported by the evidence of infections among the patients treated with tocilizumab (TCZ), for example, where studies reported the occurrence of infections as the most common side effect in patients with rheumatoid arthritis treated with TCZ [93], as well as the occurrence of herpes infections among COVID-19 patients who received high doses of glucocorticoids after TCZ administration [94]. In our review, steroids, including non-specified steroids, dexamethasone, prednisolone, and methylprednisolone, and tocilizumab contributed 18.11% and 13.86% of the CMV cohort medications, respectively—Gatto et al. additionally reported bacterial infections among COVID-19 patients who received steroids during ICU stay, also suggesting the immunomodulatory role of medications in activating virus latency [70]. However, there are conflicting evidence to whether the pharmacologic immunosuppression directly contributes to increasing the rates of infectious disease complications as impaired host immunity from a natural course of SARS-CoV-2 infection, comorbidities prior to treatment, and most importantly, administration of varying doses at different time points should also be taken into account.

Compared to a large body of literature focusing on a specific virus, we undertook a comprehensive review of all viral infections observed in COVID-19 patients, which enabled us to collect comparative clinical findings between different viruses. Our findings indicate that EBV, CMV, and HHV-6 were most either reactivated or co-occurred upon SARS-CoV-2 infection. We’ve also distinguished observed cases by reactivation and coinfection categories based on authors’ descriptions of detection criteria, which are described in Table 2, which may provide additional clarity on the status of patient’s infectious complication. Overall, evidence collected in this review indicate a potential risk of latent virus reactivation or secondary infection among COVID-19 patients and the cautious use of immunomodulatory agents. Further, high degree of similarity in symptoms between the single SARS-CoV-2 infection and coinfections with other pathogens, especially those of IAV and IBV, might imply that timely detection of reactivated and coinfected viruses is necessary to prevent severe illness and complications. However, based on the previous evidence of viral reactivations among the patients of other critical conditions such as cancer, longer ICU stay, severe pneumonia, or sepsis, imply more of a casual than absolute association between COVID-19 and detection of latent viral DNA. Thus, it is necessary to elaborate that the COVID-19 is rather an additional mechanism to other clinical conditions that may trigger viral reactivations and co-infections [95,96,97,98].

There were several limitations in this study associated with the availability of information from the selected articles. First, there were challenges in differentiating between the reactivation and coinfection of viruses as some of authors interchangeably used the terms to describe an infection or did not provide accurate detection criteria used in study besides describing an infection as either reactivation or coinfection within the text or title of article. Second, demographic information of patients of our interest was not available in all studies, which limited the extent of comparisons between different virus groups. Some cohort studies, for example, did not provide the sex ratio of population or provided the sex ratio of the entire cohort including the patients with and without reactivations or coinfections but not the target of our interest. Out of 50 extracted articles, only 9 studies provided ethnicity of patients (not shown). Third, considering the longevity of viral IgG presence upon infection that could range from several years to lifelong time upon infection [99,100,101,102], diagnostic approach with IgG detection alone in some of the studies in review may had overestimated the occurrences of viral reactivation. Lastly, reference ranges for laboratory findings either varied by studies or were not available, although the thresholds for high, normal, and low levels of each biomarker overlapped.

Conclusion

The study showed that patients were reported to have risk factors of severe COVID-19 including cardiovascular disease, diabetes, and immunosuppression as comorbidities, as well as additionally disease aggravating conditions including acute kidney injury as complication, lymphopenia and elevated D-dimer and CRP levels from laboratory testing, and steroids and antivirals for treatment [103,104,105,106,107]. Further, EBV, HHV-1, and CMV were most frequently observed within the reactivation cohort, IAV and EBV within the coinfection cohort, and EBV within the indistinguishable infection cohort. However, it is important to note that our study did not attempt to imply an increased risk of infection among individuals with latent viruses. Through this systematic review, we aimed to collect and summarize the available evidence to fulfill the literature gap on the clinical outcomes and potential risk factors of both reactivations and coinfections. Further, our review highlights a significant need for standardizing the detection criteria for reactivation and coinfection, especially within the context of comparing cases across different study sites. With the complex immune pathways triggered by SARS-CoV-2 exposing the patients to develop more adverse outcomes from other concurrent or latent infections, it is imperative that more scientific evidence become available for more efficient treatment and diagnostic measures.