Resistant Pathogens, Fungi, and Viruses

Abstract

The first reports of antibiotic pathogens occurred a few short years after the introduction of these powerful new agents, heralding a new kind of war between medicine and pathogens. Although originally described in Staphylococcus aureus, resistance among bacteria has now become a grim race to determine which classes of bacteria will become more resistant, pitting the Gram positive staphylococci, enterococci, and streptococci against the increasingly resistant Gram negative pathogens, e. g., carbapenemase-resistant enterobacteriaceae. In addition, the availability of antibacterial agents has allowed the development of whole new kinds of diseases caused by non-bacterial pathogens, related largely to fungi that are inherently resistant to antibacterials. All of these organisms are becoming more prevalent and, ultimately, more clinically relevant for surgeons.

It is ironic that despite their ubiquity in our communities, there is seldom a second thought given to viral infections in patients with surgical illness. The extent of most surgeon’s interest in viral infections ends with hepatitis and HIV, no doubt related to transmissibility as well as the implications that these viruses might have in a patient’s hepatic or immune functions. There are chapters and even textbooks written about these viruses so these will not be considered here. Instead, we will present the growing body of knowledge of the herpes family viruses and their occurrence and consequences in patients with concomitant surgical disease or critical illness. We have also chosen to focus this chapter on previously immune competent patients, as the impact of herpes family viruses in immunosuppressed patients such as transplant or AIDS patients has received thorough treatment elsewhere.

Resistant pathogens and fungi

Introduction

Infections of all kinds are an unfortunate and common condition among the surgical population. Management of these infections often requires multiple treatment modalities but usually involves some form of antimicrobial therapy. Over the years, antimicrobial resistance has become increasingly common across a wide range of pathogens. In this section, we discuss some of the more common resistant pathogens that the surgeon is likely to encounter.

HA-MRSA: Hospital-associated Methicillin-resistant Staphylococcus aureus

Staphylococcus aureus is the most commonly isolated bacterial pathogen (1). It is therefore not surprising that MRSA (both community-acquired [CA-MRSA] and hospital-acquired [HA-MRSA]) is one of the most common resistant pathogens encountered by the surgeon. Since its discovery in 1960, the incidence of nosocomial infection caused by MRSA has increased steadily. Today, S. aureus is the predominate isolate in intensive care units in the United States (2). Resistance to methicillin, and other beta lactams, results from the acquisition of the mecA gene cassette, which modifies the penicillin binding protein in the cell wall (3).

Risk Factors for HA-MRSA: (1, 4)

• Any of the following within the last year:

  • Hospitalization

  • Surgery

  • Intubation

  • Dialysis

  • Residence in long-term care facility

• Indwelling catheter or other percutaneous device

• Prior exposure to antibiotics

• Prior history of MRSA infection

• Pressure ulcer

• Colonization

The clinical spectrum of all MRSA infections ranges from asymptomatic colonization to severe invasive disease. Compared to CA-MRSA, HA-MRSA is less likely to cause skin and soft tissue infections (SSTI). SSTI accounts for only 37% of HA-MRSA infections (4). Uncomplicated abscesses 5 cm or less in diameter may be managed with incision and drainage alone. However, systemic signs or evidence of invasive disease such as cellulitis, pneumonia, endocarditis, bone or joint infection require systemic antibiotics. Vancomycin is the empiric treatment of choice where MRSA is suspected. Other agents such as linezolid, daptomycin, quinpristin/dalfopristin, and tigecycline may be considered as well (4). Local susceptibility patterns should be reviewed when choosing the appropriate antibiotic. A recent study surprisingly demonstrated no difference in outcomes between hospital and community-associated MRSA infections (5).

While the prevalence of MRSA colonization varies widely in the literature, it is an important risk factor for subsequent clinical MRSA infection. (6–9). In a recent study, 40% of all patients with a clinical MRSA infection were known to be previously colonized within the prior one-year period (9). Efforts to decolonize patients of MRSA, particularly upon admission or before surgery, have met with some success (10–15). These strategies typically consist of chlorhexidine bathing or intra-nasal mupirocin either alone or in combination. A recent multicenter randomized trial demonstrated a marked decline in clinical MRSA infections after the implementation of a universal decolonization strategy for ICU patients (15).

CA-MRSA: Community-associated Methicillin-resistant Staphylococcus aureus

In the early 1990’s, reports began to emerge of infections caused by MRSA in populations that did not exhibit traditional risk factors for MRSA colonization or infection. Typically, these patients were younger and healthier than the usual population susceptible to MRSA (1). Eventually, these infections were identified as a new strain of MRSA dubbed community-associated MRSA (CA-MRSA). CA-MRSA varies molecularly from HA-MRSA by having a smaller mec chromosomal cassette (Type IV or V compared to I, II, or III for HA-MRSA) (1, 16, 17). In the United States, USA300 and USA400 are the dominant clonal isolates, with USA300 being the most common (18). Despite its name, CA-MRSA is often encountered in the hospital setting. One recent study demonstrated that 52% of all MRSA isolates from the intensive care unit were CA-MRSA (19).

CA-MRSA has become an increasingly common pathogen, however, evaluating its true epidemiology is difficult given inconsistencies in the definition (20). The CDC definition underestimates the proportion of CA-MRSA in the population. Genetic testing is not routinely performed and is therefore not practical in defining CA-MRSA for the average physician. Others advocate a practical definition based on either temporal patterns or antimicrobial susceptibility (1). We suggest using all of these factors in evaluating patients for potential CA-MRSA infection.

Various Definitions of CA-MRSA: (1)

CDC Definition

• Outpatient diagnosis

• Diagnosis within 48 hours of admission if no other risk factors for HA-MRSA (see HA-MRSA)

Temporal Definition

• Outpatient diagnosis

• Diagnosis within 48 hours of admission

Antimicrobial Susceptibility Definition

• No or limited resistance to non-beta lactam antimicrobials (particularly clindamycin)

Risk Factors for CA-MRSA Infection: (1, 4, 9)

• Children (Neonates in particular)

• Adults age 65 or older

• Women (pregnant and post-partum)

• Athletes

• Household contacts of MRSA SSTI patients

• Emergency department patients

• Urban and/or low socioeconomic status

• Indigenous populations

• Populations living in close proximity (military, jail or prison)

• Cystic fibrosis patients

• Men who have sex with men (MSM)

• HIV patients

• Veterinarians, livestock handlers, and pet owners

• History of endocarditis

• Antibiotic exposure within the last year

• Chronic skin disorder

• Tobacco use

• Tattoo recipients

SSTI represent 90% of CA-MRSA infections (21). Typically, these infections present as a superficial abscess often mistaken for a spider bite. The presence of an abscess with surrounding erythema with a central black eschar is 94% predictive of some form of MRSA isolate. Unfortunately, CA-MRSA cannot be distinguished from HA-MRSA, MSSA, or other causes of SSTI on physical characteristics alone (1). However, one study of 137 patients presenting with cellulitis identified the presence of abscesses (OR 2.7; 95% CI, 1.3–5.8) and a body mass index (BMI) greater than 30 (OR 2.3; 95% CI, 1.1–5.0) to be independently associated with the presence of CA-MRSA.

Uncomplicated SSTI, presenting as an abscess, (without systemic signs) may be managed with incision and drainage alone (1, 17). If antibiotics are indicated (cellulitis), clindamycin or trimethoprim-sulfamethoxazole (TMP-SMX) are the empiric antibiotics of choice, since USA300 is typically sensitive to these antimicrobials (17, 18). However, as clindamycin use has increased, so has clindamycin resistance (17). Doxycycline and minocycline may also be considered. Linezolid is also an effective choice but is limited by its high cost (1). Resistance to fluouroquinolones (particularly ciprofloxacin) may be high in certain populations (MSM) (4, 18). Caution should be urged when choosing antimicrobials for cellulitis as doxycycline and TMP-SMX may not be effective against group A streptococci (GAS) (17). Local susceptibility patterns should always be considered when selecting appropriate antimicrobials.

Invasive infections, particularly pneumonia, can be rapidly fatal (mortality 50–63%). Patients with necrotizing CA-MRSA pneumonia present with hemoptysis, leukopenia, high fever and cavitary lung lesions (22). These patients have an odds ratio of dying of 11.3 (95% CI, 5.6 to 23) compared to other severe CA-MRSA infections (1, 23). One study reported a 56% mortality rate for CA-MRSA pneumonia with a median age of 14.5 (1, 24).

Regardless of the site of infection, vancomycin remains the mainstay of severe MRSA related infections. Linezolid may be considered in cases of a high vancomycin MIC’s, while clindamycin may be considered (for CA-MRSA only) as an adjunctive antimicrobial to reduce toxin production. Linezolid may also be considered in necrotizing CA-MRSA pneumonia given the relatively low lung penetration of vancomycin (1).

VRE: Vancomycin-resistant Enterococcus spp

Enterococcus faecalis and Enterococcus faecium are normal part of human intestinal flora. Combined, these account for the majority of Enterococcus infections in humans. These pathogens are notoriously difficult to treat, as they are intrinsically resistant to most penicillins, cephalosporins, and TMP-SMX. Furthermore, they easily acquire resistance to many other antibiotic classes. While some strains of E. faecalis may be susceptible to some penicillins, cephalosporins and fluoroquinolones, this is not the case for E. faecium. Resistance to vancomycin is mediated through the acquisition of a group of genes collectively known as the van gene complex. These genes encode an alteration in the cell wall with reduced affinity for vancomycin (25).

Risk factors for VRE Infection: (25, 26)

• VRE colonization – generally required for infection

• Advanced age

• Severe underlying illness

• Inter-hospital transfer

• Resident of long-term care facility

• Nutritional support (TPN)

• Central venous catheterization

• Hematologic malignancies

• Surgery for inflammatory bowel disease

• Biliary tract or liver pathology

• Transplant patients

• Hemodialysis

• Previous antibiotic exposure. Particularly:

  • Vancomycin

  • 3^rd generation cephalosporins

  • Anti-anaerobic antibiotics (Metronidazole)

  • Antibiotic combinations

  • Long-term antibiotic use

Colonized patients develop VRE infections that are similar in scope to vancomycin-susceptible isolates: intra-abdominal, skin and soft tissue, urinary tract, bloodstream and endocarditis. VRE pneumonia or CNS infections are rare (25). Approximately 8% of colonized patients develop a VRE infection either during or shortly after hospital admission (26). The associated mortality for these infections remains high (13–46%) (27).

Linezolid or daptomycin are the drugs of choice for most VRE infections. Daptomycin’s rapid bactericidal activity makes it the preferred agent in bloodstream infections and endocarditis. Some strains of E. faecalis may be susceptible to ampicillin or piperacillin, but this is becoming uncommon. Quinpristin/dalfopristin has some use against E. faecium only (25, 28). Resistance to linezolid has been reported and is an emerging problem (28).

Susceptible patients while undergoing medical care easily acquire VRE. A recent study found that 12.3% of patients who were initially VRE negative were colonized before leaving the intensive care unit (29). VRE transmission often occurs via healthcare workers and once acquired may be life-long. Therefore methods to reduce transmission such as active screening and isolation have been instituted in some locations. Screening and isolation for VRE-positive patients, however, remains controversial with no consensus criteria for the removal of patients from isolation (25).

ESBL: Extended-spectrum Beta-lactamase Producing Bacteria

Many organisms have acquired, either through point mutation or plasmid acquisition, the ability to produce a group of enzymes collectively known as extended-spectrum beta-lactamases. Carapenem-resistant Enterobacteriaceae (discussed below) actually represent a special case of this phenomenon. These enzymes were originally described as a point mutation in the classic TEM and SHV beta-lactamases, thereby conferring expanded activity. Soon other plasmid mediated ESBL enzymes were also discovered. Unlike CRE’s, ESBL enzymes are not limited to the Enterobacteriaceae family and are commonly found in Pseudomonas species. In 2007, a survey identified that up to 17% of K. pneumoniae and 10% of E. coli demonstrated ESBL activity. Most of these (65%) produce members of the CTX-M class of beta-lactamases (30).

ESBL transmission between pathogens generally occurs via a large plasmid, which often encodes resistance to other antibiotic classes, particularly fluoroquinolones, aminoglycosides and TMP-SMX. The high likelihood of concomitant resistance among ESBL producing pathogens limits therapeutic options. For this reason cephalosporins, fluoroquinolones, and TMP-SMX are not appropriate options for treating most ESBL infections. Unless CRE is suspected, carbapenems remain the antibiotic class of choice for treating these infections (30, 31). Although the data is limited, beta-lactam/beta-lactamase inhibitor combinations, such as piperacillin/tazobactam, have also demonstrated efficacy in treating these infections, particularly in the urinary tract (30–32). Among the beta-lactamase inhibitors, tazobactam is the most potent and is active against many ESBL classes (TEM, SHV, and CTX-M) (30). Tigecycline, colistin, and fosfomycin are additional therapeutic options (30, 32). The wide range of ESBL activities, along with associated resistances, highlight the need to perform appropriate antibiotic sensitivity assays, as well as consult local antibiotic susceptibility patterns when choosing the appropriate therapeutic agent.

CRE: Carbapenem-resistant Enterobacteriaceae

Enterobacteriaceae is a large family of gram-negative rods that contains many common human pathogens, including: Escherischia coli, Klebsiella species, Enterobacter species, and over 70 other genera. Resistance to many broad-spectrum antibiotics is common among members of this family and until recently physicians could depend on the carbapenem antimicrobial class to reliably treat these pathogens. However, since 2000, carbapenem resistance has been growing. While still uncommon, the percentage of Enterobacteriaceae in the United States that were CRE increased from 1.2% in 2001 to 4.2% in 2011. The largest increase was among Klebsiella species, particularly K. pneumoniae (33). Another report describes carbapenem resistance in E. coli at 4.0% and at 10.8% among K. pneumoniae (34).

CRE produce ESBL with the largest spectrum of activity. Unlike MRSA or VRE, carbapenem resistance is not mediated by a single set of genes within a single species. Instead, CRE is mediated through multiple plasmid-encoded enzymes across an entire family of organisms (33, 35). The most common resistance gene is the highly transmissible Klebsiella pneumoniae carbapenemase (kpc), so named because of its initial discovery in that species (33). The metallo-beta-lactamases (MBLs) VIM (Verona integrin-encoded MBL) and IMPs (active on imipenem) are also common (34).

Risk Factors for CRE: (35, 36)

• Advanced age

• Intensive care unit stay in previous 3 months

• Central venous catheterization

• Receipt of antibiotics (particularly fluoroquinolones) in previous 3 months

• Diabetes mellitus

• Recent invasive procedure in the last 6 months

• Isolation of resistant bacteria in previous 6 months

• Dependent functional status

• Permanent residency in institution

• Charlson comorbidity index greater or equal to 3

CRE infection and colonization have been managed with strict cohorting and isolation (37). A recent study suggests that, unlike VRE, asymptomatic colonization may not be life-long. This study found that at one year following the index admission with CRE, only 39% of patients remained positive. A lack of hospitalization during this time increased the likelihood of becoming CRE negative (38).

CRE infections carry a high mortality rate between 40–50%, which is increased to 72% in CRE bloodstream infections (33, 39). The current mainstays of treatment include colistin, tigecycline, and aminoglycosides. Fosfomycin and polymixin are additional options. Some authors advocate prolonged profusion carbapenem or double carbapenem therapy for CRE with an MIC less than or equal to 4 mg/L (39).

Candida

Candida species are the most common invasive fungal pathogens in humans. It is the third most common cause of infection overall and is the second most common pathogen in North American ICU’s (40). While Candida albicans remains the most common individual isolate in the United States and Canada, non-albicans species make up 57.1% of all cases of candidemia (40). When all forms of candidiasis are considered, including hair, skin, and nail infections, C. albicans is the causative pathogen in 80% of cases (41). The incidence of invasive candidiasis is increasing. Between 2003 and 2005, the incidence of candidemia increased from 3.65 to 5.56 per 100,000 people (42).

Risk Factors for Invasive Candidiasis: (40–42)

• Colonization of several body sites

• Extremes of age

• Exposure to broad spectrum antibiotics

• Immunocompromised

  • Cytotoxic chemotherapy

  • Corticosteroids

  • Transplant

  • Neutropenia

  • HIV

• Disruption of the physiological barriers of the GI tract

• Major abdominal surgery

• Other surgery during hospitalization

• Surgery on the urinary system in the setting of candiduria

• Major trauma (ISS > 20)

• APACHE II score > 20

• Candiduria > 10⁵ cfu/ml

• Diabetes

• Hemodialysis

• Mechanical ventilation

• Central venous catheterization

• Enterocutaneous fistula

• Total parenteral nutrition

• ICU stay > 7 days

• Multiple transfusions

Colonization by Candida species is a key risk factor for invasive candidiasis. Alteration of normal host flora via broad-spectrum antibiotic exposure allows for fungal overgrowth. Increasing burden of Candida as demonstrated by semiquantitative cultures from multiple sites at multiple time points has been associated with the development of invasive candidiasis. Some view the identification of Candida from more than two body sites as a justification for antifungal therapy (42). Eggimann, et al. describes criteria for “pre-emptive” antifungal therapy as substantial colonization in the presence of multiple risk factors. Likewise, “prophylactic” antifungal therapy is justified for certain subgroups at particularly high risk for infection: organ-transplant recipients or immunocompromised patients with expected or long-term neutropenia (43).

Selective pressure from the frequent use of prophylactic fluconazole has contributed to the increase in azole-resistant Candida species, particularly C. glabrata and C. krusei (40, 42, 44). C. glabrata is the most common of the non-albicans species and tends to occur in patients with prior antifungal therapy, older patients, and transplant recipients (both solid organ and hematopoietic stem cell transplants). It is uncommon in younger patients and neonates (40, 41). Risk factors for C. parapsilosis infection include recent surgery, younger age, transplant patients, and those receiving TPN. It is also a frequent NICU pathogen (40, 41). C. Tropicalis is common in patients with hematologic malignancies and neutropenia (40, 41). Most patients with C. krusei candidemia have had prior antifungal exposure, neutropenic or received a hematopoietic stem cell transplant (40).

Overall, mortality following fungal infection remains high but varies somewhat based upon the individual pathogen. Ninety-day mortality following candidemia ranged from 30% for C. parapsilosis to 46.4% for C. krusei (40). Overall mortality for invasive candidiasis ranges between 40–60% but can be as high as 80% in selected populations (immunocompromised) (42, 43)

Azole class antifungals, particularly fluconazole, are the most commonly prescribed antifungals in the surgical population. This usage is probably appropriate in many cases, particularly for proven C. albicans infection or empirically in an otherwise low-risk patient. However, in critically ill patients or those with previous azole exposure, echinocandins (caspofungin, micafungin, anidulafungin) should be considered first-line agents. Amphotericin B is another option for azole-resistant strains but comes with a significant side effect profile. Lipid formulations have lowered but not eliminated the risks of nephrotoxicity and infusion-related reactions. Therefore, amphotericin B should be reserved for salvage therapy (45).

Viruses

Human Herpes Viruses in Surgical Patients

Introduction of human herpes viruses

There are 8 human herpes viruses (HHV), including herpes simplex virus – 1 (HSV-1/HHV-1), herpes simplex – 2 (HSV-2/HHV-2), varicella-zoster virus (VZV/HHV-3), Epstein-Barr virus (EBV/HHV-4), cytomegalovirus (CMV/HHV-5), Roseola virus (HHV-6 and HHV-7), and Kaposi Sarcoma associated virus (KSHV/HHV-8). As summarized in the table below, these viruses are highly prevalent in humans (46–51). These viruses share a similar life cycle of primary infection, often inducing mononucleosis or flu-like symptoms, followed by control and disease resolution in immune competent hosts. Following resolution, these viruses then enter a state of latency, characterized as relative dormancy with little if any appreciable viral replication during most of the host’s life. Although these viruses are known to reactivate in immunosuppressed individuals, sometimes causing devastating disease, until recently little attention has been given to these viruses in immune competent hosts. During the last three decades, there has been increasing awareness of reactivation of latent herpes viruses in immune competent hosts with surgical disease.

Overview and Epidemiology of Herpes Viruses (Table 1)

Table 1.

Overview and Epidemiology of Herpes Viruses

Common Name	Formal Name	Prevalence	Primary Targets	Transmission	Sites of Latency	Primary Disease
Herpes Simplex Virus-1	HHV1	50–80%	mucoepithelia	oropharyngeal contact	Sensory and cranial nerve ganglia	cold sores
Herpes Simplex Virus- 2	HHV 2	20–25%	mucoepithelia	sexual contact, congenital	Sensory nerve ganglia	genital lesions
Varicella Zoster Virus	HHV 3	>90% (prevaccine)	mucoepithelia	airborne	dorsal root ganglia	chicken pox
Epstein-Barr Virus	HHV 4	95%	Epithelial, oral lymphoid cells	saliva	B lymphocytes	infectious mononucleosis
Cytomegalovirus	HHV 5	50–80%	Monocytes, lymphocytes, and epithelia	saliva, sexual contact, congenital, blood transfusions, transplant	Monocytes, Lymphocytes	asymptomatic, mono-like symptoms
Roseola Virus	HHV 6, 7	100%	T cells	saliva	various leukocytes	asymptomatic (90%), roseola infantum (10%)
Kaposi Sarcoma associated Virus	HHV 8	varies	Lymphocytes and epithelia	sexual contact, saliva	B cells	asymptomatic, oncogenic in immune-suppressed

Latency and reactivation

To better understand reactivation, it is important to first define latency. In general, herpes viruses become quiescent after infection, with viral DNA present in host cells with very little (if any) transcriptional activity. Such latency seems to be maintained by a combination of host immunity and epigenetic regulation. Following primary infections, the host mounts concomitant innate and adaptive immune responses to control viral spread, but control typically occurs well after full dissemination of virus to its target cells/tissues. Thus, in addition to inducing non-specific innate anti-viral responses, all HHV induce epitope specific immunoglobulin and CD4/CD8 T-cell responses. It is important to note that most of what we know about latency and reactivation mechanisms comes from animal models of CMV and HSV, with less known about VZV, EBV, or HHV6-8 because of a lack of good animal models for these infections.

The importance of epigenetics in herpesvirus latency is becoming increasingly clear. Herpesvirus DNA are not typically integrated into host DNA, but are maintained as episomes in infected cells. Like other eukaryotic DNA, HHV-DNA become wrapped around histone proteins in a repeating nucleosome pattern (reviewed in (52)). This leaves most of the viral genomes inaccessible to transcription and replication. It is clear that most viral genomes in host tissues are thereby epigenetically regulated and quiescent at any given time (53–55){Liu, 2008 #1661}. This “chromatinization” of viral DNA occurs very rapidly after infection, thereby contributing to development of latency (56).

Most likely as a viral survival advantage, this epigenetic regulation can be interrupted, leading to localized reactivation events (57). Such reactivation events likely lead to transient viral replication and shedding, thus allowing perpetuation of virus within communities. In immune competent hosts, however, these reactivation episodes are quickly controlled by memory T and B cell responses, leading to resumption of latency. In contrast, hosts with impaired immunity often have reactivation episodes that progress to viral disease, with shedding and transmission of live virus. Unlike HHV with cutaneous manifestations, such as HSV-1/2 and VZV, whose reactivation episodes are obvious, it is unknown how frequently the other HHV may reactivate in immune competent hosts. It does seem clear, however, that transient compromise in host immunity will allow transcriptional reactivation of these viruses (58, 59). For the purposes of this article, we will define immune competent hosts as those not undergoing canonical immune suppression or have disease related immune compromise (such as AIDS), understanding that following surgery, trauma or critical illness there may very well be transient compromise in host immunity (60).

Diagnosis of HHV infection/reactivation

One of the major obstacles in our understanding of HHV infection and reactivation is that with the exception of HSV and VZV, there are no cutaneous manifestations of reactivation. For most, it is hard not to notice a perioral cold sore, or the painful outbreak of VZV/HHV-3 in the form of herpes zoster, making diagnosis of reactivation episodes for these viruses a simple matter. The other herpes viruses do not typically have cutaneous manifestations, thus requiring serologic or tissue diagnosis. Prior to introduction of the “monospot test”, Paul and Bunnels testing for EBV infection was the method used to detect EBV associated infectious mononucleosis (61). Although immune globulin monitoring remains one of the best ways to diagnose acute or previous HHV infection, for monitoring/diagnosing reactivation it has far less accuracy and utility. Immune globulin monitoring has therefore been mostly replaced by DNA based molecular methods to diagnose HHV reactivation in immune competent hosts.

Triggers of HHV reactivation

There are myriad triggers for reactivation of latent HHV, and one of the best studied in healthy hosts is stress. Glaser et al. were among the first to show that the social stress during academic examinations can induce reactivation of HHV in healthy medical students (62). These findings were supported further by an animal model of HSV-1/HHV-1 reactivation following social stress (63). Certainly among the most healthy immune competent hosts are astronauts, and there are numerous studies that show the preflight stress as well as the stress of space flight can stimulate reactivation of CMV, HSV, VZV and EBV (64–67). It should therefore come as no surprise in the sections that follow that more aggressive stressors, such as surgical disease or infections that induce critical illness are also triggers for viral reactivation.

HHV following Cardiac Surgery

Herpesviruses were first associated with surgical disease in immune competent patients in the late 50’s. Battele and Hewlett, and subsequently others, described a peculiar viral illness that befell 4–30% of patients undergoing cardiac surgery with extracorporeal bypass (68–73). These cases usually occurred 3–6 weeks after surgery, had symptoms consistent with viral mononucleosis, but most were EBV-negative (by Paul Bunnell testing). Ultimately, Paloheimo et al made the first connection between these febrile illnesses and CMV, and it was later concluded that many cases were likely a consequence of blood transfusion practices during extracorporeal circulation in that era (74). Because of limitations in diagnostics at that time, it remained unclear whether post-pump CMV was a primary infection or reactivation of latent virus until later work suggested that most were reactivations (75).

The consequences of CMV reactivation in cardiac surgery patients remain unclear. Although the early observations of CMV activity were not linked with worsened outcomes, later work in patients with complications after cardiac surgery suggested otherwise (76). It was observed that patients with CMV infection concomitant to bacterial infections often had far worse outcomes (77). For example, in patients with mediastinitis following cardiac surgery, CMV viremia or viruria was associated with higher mortality and impaired clearance of local infection (76). This was hypothesized to be related to impaired neutrophil function (78). Likewise, Rand observed that mice given primary CMV plus bacterial infection had experienced 80–100% mortality, compared to none in control groups (79).

Less is known about EBV and HSV associated disease in cardiac surgery. VZV has been reported to reactivate as herpes zoster following cardiac surgeries (80–82). Prompt diagnosis and treatment is important as misdiagnosis can lead to short-term and long-term pain control issues or other related sequelae.

HHV have also been associated with other cardiac diseases. CMV viremia correlates with disease severity in patients admitted with acute heart failure (p < 0.001) and this is thought to be secondary to the pro-inflammatory state associated with heart failure progression. (83). HHV have also been implicated in atherosclerosis, and biopsy of the major arteries in young trauma patients identified HSV and CMV in atherosclerotic lesions and foam cells of the intimal layer (84, 85). Acute infection with CMV in rats causes endothelial injury (86) and some have postulated that this contributes to hypertension (87). Later studies evaluating mechanisms indicate that presence of lifelong, latent herpesviruses in atherosclerotic plaque may exert pathogenic effects by penetrating the arterial wall, modifying lipid metabolism, and stimulating the production of pro-inflammatory cytokines and growth factors (88).

HHV after trauma/burn injury

Although HHVs have likely been reactivating in humans since their acquaintance hundreds of thousands of years ago (89), description of reactivation events didn’t become common until the 20th century. It is perhaps not surprising that the first described trauma associated HHV reactivation was a case of VZV/HHV-3 herpes zoster, although its association with a lightning strike makes it a bit remarkable (90). The association between herpes zoster and trauma has been subsequently elaborated in case reports and series, and recently confirmed in a large case control study of Medicare patients (91). The first report of CMV/HHV-5 related to trauma followed some years after recognition of CMV reactivation in cardiac surgery patients, and occurred in a man who died of disseminated CMV disease following severe facial fractures (92). Soon after this, the first case of HSV/HHV-1 was reported after oral trauma (93). It was several years later that the first confirmed case of EBV/HHV-4 was reported in a patient that had suffered trauma requiring splenectomy and transfusions (94).

HHVs also have a long-standing association with burn injury, beginning with the observations of systemic CMV (95) and herpes simplex activity in burn wounds (95). Some years later other reports emerged describing CMV infections in patients with burn injuries (96–98). Subsequent studies suggested high rates of reactivation (>50%) as well as primary infections (12–15%) following burn injury (99, 100). Major burns have subsequently been shown to be one of the strongest CMV reactivation stimuli in humans with incidence as high as ~70% (101, 102). There are only case reports of VZV in adult burn patients (103), and pediatric burn patients with primary VZV have higher likelihoods for pneumonitis (104). It seems that the editors of the British Medical Journal were prescient when they observed that “…there seems to be at least as much for virologists to do in the surgical wards as in the medical ones.” (105). Unfortunately, at least for burns, little attention is being paid in the US to this entity (106).

This lack of attention may be because it is unclear what impact HHV may have on outcomes in burns. Reactivation of CMV in burn patients has not been associated with worsened mortality (98, 99, 107), although it has been associated with longer ICU stays and duration of mechanical ventilation (101). Primary infection in burn patients is also of concern, as CMV can be detected in cadaveric skin grafts (108) and transmission by seropositive skin grafts can cause severe disease (109). Cutaneous HSV reactivation, seen in up to 25% of burns, presents 1–3 weeks post-burn as a cluster of vesicles around the margins of a healing burn (110). Areas of active epidermal regeneration appear to be the most commonly affected (100), but progression to visceral HSV involvement has been described, manifesting as necrotizing adrenal and hepatic lesions (95). HSV reactivation of the respiratory tract is seen in up to 50% of burn patients (110, 111). The impact of pulmonary HSV reactivation on mortality has been mixed, with larger prospective studies still needed (107, 112, 113)).

HHV in Critical Care/Sepsis

Similar to trauma patients, there have been numerous reports of HHV eruptions/cutaneous manifestations during critical illness, but those HHVs without obvious outward signs have taken much longer to recognize as an entity. During the mid to late nineties, several different investigators began reporting herpesvirus infections/reactivations in previously immune competent patients during critical illness. These first reports included both HSV and CMV, and since then more limited work has evaluated the other herpes family viruses during critical illness. There is a growing body of knowledge in this area that includes both medical and surgical patients, and the use of animal models has contributed significantly to our understanding of mechanisms and consequences of such reactivation events. Although it has not been confirmed for all HHV, because of the similarities in infection, control and development of latency between the HHVs, it seems likely that each HHV follows a pattern of reactivation similar to CMV. In the sections that follow, we will review each of the herpes family viruses and their associations with critical illness in previously immune competent patients.

It has become clear that CMV can reactivate during critical illness, with sensitive PCR based methods showing that approximately one-third of latently infected patients have CMV reactivation during critical illness (114). These reactivation events can lead to live virus shed in the blood and pulmonary secretions of affected hosts (115, 116). It seems likely that these reactivation events are consequent to inflammatory insults that trigger release of cytokines, epigenetic deregulation of viral genomes, and possibly immune compromise (58, 67, 117–119). It remains unclear, however, whether CMV is a true pathogen or merely an indicator of severity of illness. When considering all of the currently available studies (76, 101, 115, 116, 120–129), reactivation is associated with roughly doubled risks of hospital mortality and duration of mechanical ventilation (Figure 1). When one considers the potential pathogenic mechanisms associated with CMV, including pulmonary injury (117) and immune modulation (130), the possibility that CMV actually contributes to poor outcomes in immune competent patients is intriguing.

Figure 1.

HSV-1/HHV1 is the second best studied HHV that reactivates during critical illness. In mechanically ventilated patients, HSV-1 can be detected in tracheal aspirates and in lower respiratory tract secretions of intensive care unit patients in 22–54% and 16–32% of cases, respectively (131). This is confused however by the observation that asymptomatic shedding of HSV can occur in up to 5–10% of healthy individuals (132, 133). Furthermore, HSV haplotypes isolated from lower respiratory tracts of intubated patients have been shown to be identical to those isolated from the oropharynx, suggesting possible spread down the tracheobronchial tree through secretions (134). The lack of association of HSV reactivation with worsened outcomes in numerous studies casts further doubt on its importance as a possible pathogen (116, 128, 135–138). Nonetheless there are reports that link HSV reactivation with prolonged mechanical ventilation (136, 138), prolonged ICU stays (136), and even increased mortality (131, 139, 140). One possible explanation is that viral load is important as suggested by worsened mortality in patients with high viral loads (141). For now, however, the preponderance of data suggests that HSV activity during critical illness is simply an indicator of disease severity.

To date there are very few data for the other HHV in critically ill patients. EBV has been thought to reactivate during times of stress as detected by elevated antibody levels, which has since been confirmed by data showing EBV reactivation in 61% of ICU patients (142, 143). The lack of in-vivo models has impeded progress in understanding EBV reactivation, and therefore there are no published data on mechanisms or consequences of EBV reactivation in immune competent hosts. Similarly there are very few data for VZV/HHV3 and HHV6-8 during critical illness. VZV has been reported following spinal surgery but otherwise remains understudied in critically ill patients (81, 144–147). As with other HHV, aberrations in cell-mediated immunity are thought to be one of the causes of VZV reactivation (148, 149). HHV6 and HHV7 have been shown to reactivate during critical illness, but the consequence of this is unknown (150). Finally, like CMV, KSHV/HHV8 has been associated with lung disease (idiopathic pulmonary fibrosis), but has not yet been reported in critically ill immune competent hosts (151).

HHV in Gastrointestinal Disease

HHV can cause a variety of gastrointestinal diseases, many of which have surgical implications. One example is the relatively long-standing relationship between HHV and gastrointestinal ulcerative diseases in immune suppressed HIV and transplant patients. It is now recognized that immune competent trauma, critical care and postoperative patients can also suffer from HHV-related intestinal ulceration. CMV for example can cause colonic mucosal ulceration sometimes leading to perforation in immune competent hosts (152–154). Because this presentation can mimic other infectious colitides, CMV colitis should be considered when an infectious etiology is suspected but cannot be identified. Severe inflammation and ulcerative lesions have primarily been noted in areas of predominant endothelial distribution of CMV inclusion bodies (155), and ischemia from narrowing of the capillary circulation has been postulated in ulcer/perforation pathogenesis. CMV enteritis has been suggested to have higher mortality in immune competent patients than immunocompromised populations, possibly due to lower index of suspicion (156, 157). Although CMV is a common cause of upper GI ulceration in transplant and AIDS patients, it does not appear to play a meaningful role in immune competent patients (158). Diagnosis of HHV associated ulceration can be made with endoscopic or surgical biopsy, but is often made in surgical specimens post-hoc. Whether patients with perforation and subsequent diagnoses made by pathologic evaluation will benefit from antiviral treatment is unclear and will require further study.

Another common disease recently associated with HHV is appendicitis. There has been a long standing suspicion that non-HHV viral infections are associated with appendicitis, given it’s bimodal and seasonal occurrence (158). The first described association of HHV with appendicitis was in an AIDS patient (159), but a recent study of childhood appendicitis has shown periodic reactivation (21%) of CMV in lymphoid tissue of the appendix (160). In this study, HHV-6 was also identified in about 8% of specimens. It is unclear whether this is a cause or simply a consequence of the appendicitis, given that sepsis can cause HHV reactivation (161).

There is also a long-standing relationship between HHV/infectious mononucleosis and splenic disease. HHV have been known for many years to cause splenomegaly, which on occasion can lead to splenic rupture (162). Spontaneous splenic rupture is less common in acute CMV infection than EBV infection, despite one-third of acute CMV infections demonstrating splenomegaly (163–165). In contrast, HHV infection after splenectomy likely represents a distinct clinicopathologic syndrome (166). Acute CMV infection was first identified in post-splenectomy trauma patients in 1982 (167). Review of case reports shows that these infections typically occur within 2 to 4 weeks after splenectomy (166–169). The syndrome likely results from poor control of early viremia because of the lack of both splenic function and the typical brisk IgM response. Although it is difficult to determine acute versus reactivated CMV in these cases one study using anti-CMV IgG maturation indices supports acute infection in this syndrome (170). These cases of widely disseminated post-splenectomy CMV can sometimes be fatal (153).

Treatment of HHV reactivation/infection

There are few data to support or refute treatment of HHV reactivation in surgical patients. While there are scattered reports of treatment in immune competent patients, these all follow reactivation, likely suffer from selection bias and perhaps not surprisingly show no benefit. The best available data come from animal models, showing that CMV reactivation events induced by sepsis can be prevented with antiviral therapy (117, 171). Unfortunately antiviral treatment works best if administered prophylactically, which would require treating all patients, many of whom would never develop reactivation (172). For now there are no good data to support treatment of critically ill patients with HHV reactivation outside of a clinical trial (173). Fortunately there is a randomized control trial evaluating ganciclovir for prevention of CMV reactivation and acute lung injury in immune competent hosts (ClinicalTrials.gov #NCT01335932).

There are also no strong data for treatment of other HHV reactivations in critical care populations. A randomized control trial showed acyclovir treatment can prevent HSV reactivation, but appears to have no effect on survival or duration of mechanical ventilation (174, 175). Clinically severe or symptomatic cutaneous HSV or herpes zoster reactivations are usually treated with acyclovir once diagnosis is confirmed (176). There are some concerns that delay in diagnosis and treatment of HSV may lead to systemic dissemination causing necrotizing hepatic and adrenal lesions, bacterial superinfection, and even death, although data supporting this are lacking (97, 177). There are no data to support or refute treatment of HHV 2, 4, 6, 7 or 8 during critical illness in immune competent hosts.

Conclusion

The complexity of patients managed by surgeons continues to increase. With this complexity comes the unique host susceptibility to infections with microbes that were unknown pathogens even 50 years ago, including antimicrobial-resistant bacteria, fungi, and viruses. Although most surgeons will not primarily manage these organisms, it will be important for them to maintain a working knowledge of them to be able to provide optimum care for their most vulnerable patients.

Key Points.

1. The complexity of patients managed by surgeons continues to increase. With this complexity comes the unique host susceptibility to infections with microbes that were unknown pathogens even 50 years ago, including antimicrobial-resistant bacteria, fungi, and viruses.

2. Although most surgeons will not primarily manage these organisms, it will be important for them to maintain a working knowledge of them to be able to provide optimum care for their most vulnerable patients.