Published online before print December 10, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Center for Vaccine Development, Department of Medicine, University of Maryland School of Medicine, Baltimore, Maryland, USA
1Correspondence: Center for Vaccine Development, Department of Medicine, University of Maryland School of Medicine, 685 W. Baltimore Street, HSF I, Suite 480, Baltimore, MD 21201, USA. E-mail: across{at}medicine.umaryland.edu
ABSTRACT
Immunization is a highly effective public health measure that reduces the incidence of infectious diseases, yet there has been relatively little effort toward the development of vaccines for nosocomial infections. Many nosocomial infections originate on mucosal surfaces (e.g., respiratory or gastrointestinal mucosa). As patients who are hospitalized once are more likely to be hospitalized again, we propose a prime-boost immunization strategy, whereby a priming dose of vaccine for a nosocomial infection is administered mucosally. Upon readmission, a parenteral boost would elicit a rapid immune response locally and systemically. Such a strategy could reduce or ameliorate nosocomial infections and perhaps limit dissemination of nosocomial pathogens. Thus, a more aggressive effort to develop vaccines for nosocomial infections is warranted.
Key Words: mucosal • vaccine • prime-boost
INTRODUCTION
Although vaccines are firmly established in the public health sector as an effective means of preventing infection, relatively little attention has been focused on the potential role of vaccines for the prevention/amelioration of nosocomial infections, especially for those that occur in the critically ill patient. Nosocomial infections are widely prevalent and account for a substantial increase in morbidity and mortality among hospitalized patients. In addition, bacteria that are resistant or poorly responsive to multiple antibiotics often cause these infections. In this review, it will be argued that it is possible to identify populations for immunization with nosocomial vaccines; identify and prioritize infections that might be targeted; and devise immunization strategies that maximize the likelihood of an immunologic response. This being the case, a more aggressive approach to the development of vaccines for nosocomial infections is warranted.
CONDITIONS THAT MIGHT MERIT IMMUNIZATION WERE VACCINES AVAILABLE
Patients who are hospitalized on one occasion are more likely to have subsequent hospitalizations, many of which may be complicated by nosocomial infections [1 ]. A number of infections occur on a recurring basis in hospitalized patients and include pneumonia, Clostridium difficile colitis and staphylococcal infections, as well a consequence of infection, sepsis. Patients who enter intensive care units (ICUs) are at risk of acquiring nosocomial pneumonia, often with Gram-negative bacteria. Indeed, 45% of ICU patients become colonized with Gram-negative bacilli, and 22% occur in the first 24 h [2 ]. Nosocomial pneumonia developed in 23% of colonized patients in comparison with 3.3% of those noncolonized. C. difficile colitis is an increasingly common and virulent infection caused by toxins elaborated by this anaerobic organism. Although this infection is most often associated with antibiotic use (it accounts for 15–25% of antibiotic-associated diarrhea), it also occurs in patients receiving chemotherapy [3 ]. Although it is treatable with antibiotics, there are frequent recurrences. Patients who have at least one episode of recurrence have a 50–65% risk of having an additional episode [3 ]. Staphylococcal infections are consistently a leading cause of nosocomial infections. More recently, with the increase in methicillin-resistant Staphylococcus aureus (MRSA) isolates, it has become increasingly difficult to treat with clinical progression occurring while on antibiotic therapy. Ominously, with the acquisition of the Pantin-Valentine-Leukocidin virulence factor by community-acquired MRSA, a more fulminant form of staphylococcal infection is now prevalent in the community and often requires hospitalization [4 ]. Finally, sepsis may complicate the clinical course of previously healthy patients admitted for trauma or wound infection or of patients admitted to the hospital for a wide spectrum of chronic illnesses. Once it develops, sepsis is associated with 20–50% mortality, even with optimal antibiotics and supportive care [5 ]. In all of these infections, enhancement of the immune system through immunization might reduce the acquisition of organisms, ameliorate the progression of the infection, and decrease their spread.
RATIONALE FOR MUCOSAL IMMUNIZATION AGAINST NOSOCOMIAL INFECTIONS
There are a limited number of vaccines that have been developed for nosocomial infections. Clinical trials with parenteral vaccines have been undertaken with pneumococcal vaccines in hospitalized patients, an experimental vaccine for S. aureus in hemodialysis patients, and Pseudomonas aeruginosa vaccines in patients with cystic fibrosis [6 7 8 ]. No vaccines for sepsis or C. difficile have been tested in clinical trials. Most of these vaccines have been delivered parenterally, which protects the bloodstream from disseminated infection but may have limited capacity to prevent initial infection or colonization (with the proven exception of the pneumococcal conjugate vaccine [9 ]). There is increasing interest in giving vaccines via the mucosal surface. Many pathogens inflict their effect exclusively (or mainly) along the mucosa of the gut (e.g., C. difficile), respiratory, or urogenital tracts. Thus, the induction of an immune response to Gram-positive or -negative bacteria or to toxins (e.g., diphtheria, anthrax, tetanus) at the point of entry of systemic infections may not only provide systemic protection but may also limit the initial local infection or colonization. Several excellent reviews of the mechanisms of mucosal immunity induction have been published [10 , 11 ]. Vaccines that target mucosal inductive sites (e.g., gut-associated lymphoid tissue; nasal-associated lymphoid; bronchus-associated lymphoid tissue) stimulate all arms of the immune system, including the induction of mucosal secretory IgA, functional serum IgG antibodies, and systemic and local cell-mediated immune responses [10 , 11 ]. Naturally acquired mucosal infections can provide life-long immunity, and mucosal vaccines elicit immunity for longer than 7 years [12 ]. Finally, unlike the case with parenteral vaccines, multiple mucosal vaccines may be given concomitantly or in combination.
Several vaccines have been licensed for mucosal administration. Currently, the U.S. Food and Drug Administration has approved a cold-adapted, attenuated influenza vaccine (Flumist®), which is administered by the mucosal [intranasal (i.n.)] route [13 ]. Additional influenza vaccines have been tested, and additional mucosal vaccines have been licensed for polio (Sabin oral polio vaccine), typhoid fever (Ty21a), cholera (CVD 103-HgR), and rotavirus infection (RotaTeq®).
MUCOSAL IMMUNIZATION STRATEGIES
Vaccines can be administered by the mucosal route using a number of strategies. Attenuated, live bacterial vectors (e.g., Salmonella Typhi, Shigella) can be engineered to carry foreign antigens of interest [14 ]. One can also deliver DNA vaccines mucosally [15 ]. Other investigators have used live viral vectors, such as adenovirus [16 ], which deliver vaccine to targeted cells. Vaccine antigens of interest can be delivered in nonliving delivery systems as well. Antigens can be expressed on viral-like particles; embedded in liposomes, cochleates, biodegradable microspheres, or immune-stimulating complexes; and delivered to mucosal, inductive sites. There has also been interest in developing transgenic plants that express vaccine antigens. Upon ingestion of these "edible vaccines," the host immune system develops a mucosal immune response [17 ]. A live, attenuated vaccine for tularemia (live vaccine strain) was delivered by aerosol on a massive scale to Soviet soldiers during World War II and to individual healthy subjects by Hornick [studies reviewed in ref. 18 ].
One attractive immunization strategy is to deliver an initial priming dose of vaccine by the mucosal route, and then at some later time, give a parenteral boost ("heterologous prime/boost"). The delivery of the initial dose of vaccine via the mucosal route is easy and relatively inexpensive to administer, well-tolerated, and widely accepted. In a recent study by Vindurampulle et al. [14 ], mice were given two doses of tetanus toxoid, 28 days apart. Before the second dose of vaccine, the tetanus-neutralizing antitoxin level as 0.10 international units (IU)/ml, and by Day 56, it was 0.70 IU/ml (Table 1 ). When tetanus antigen was delivered i.n. via an attenuated S. Typhi carrier, there was a similar response; however, when mice were primed i.n. with the tetanus antigen-containing S. Typhi followed by a parenteral dose of tetanus toxoid, the Day 35 response had doubled, and by Day 56, the response was significantly (P<0.01) superior to either alternative regimen. Later in this chapter, we will argue that a heterologous prime-boost strategy may be ideal for the prevention or amelioration of nosocomial infections.
|
View this table: [in a new window] |
Table 1. Mucosal Priming with S. Typhi-Based Fragment C Vaccine Allows Anamnestic Response to Parenteral Tetanus Toxoid
|
As is the case with parenteral vaccines, the immune response to mucosally delivered vaccines is enhanced with adjuvants. Escherichia coli heat-labile enterotoxin (LT) and cholera toxin (CT) are potent, mucosal adjuvants that increase the uptake of antigens by epithelial cells, increase antigen presentation by APCs, increase IgA formation, and increase TH1 and TH2 immune responses [19 , 20 ]. The unmodified toxins, however, are too reactogenic to be given as oral adjuvants. Therefore, investigators have made mutant LTs or CTs that retain adjuvanticity but have minimal toxicity. In a related approach, CTA1-DD was developed as an adjuvant, whereby the enzymatically active A subunit domain of the toxin is linked to the cell-binding domain of S. aureus protein A (see ref. [20 ]; rather than the CT B subunit); in this way, enterocytes are not targeted. Bacterial DNA with cytosine phosphate-guanosine (CpG) motifs has been administered i.n. with vaccine antigens (e.g., hepatitis B surface antigen), with and without alum [21 ]. CpG-bearing DNA stimulates immune responses through TLR9. IL-1 and RANTES also have been given experimentally as nasal adjuvants. Although not an adjuvant, chitosan (polycationic polysaccharide) has been coadministered with i.n. vaccines to retain the vaccine at the inductive site [22 , 23 ].
One cautionary consideration for mucosal adjuvants, particularly when given by the i.n. route, is that some of the mucosal adjuvants (e.g., LT, CT, and their mutants), as well as some antigens, bind to GM1 and other gangliosides on neurons. Because of this, such molecules can be transported retrogressively across the cribiform plate via the nerve fiber sheaths to the CNS. An inactivated influenza virosome adjuvanted with LT (Nasalflu®) was withdrawn from the market after a 19-fold increase in the incidence of Bell’s palsy was reported following i.n. immunization [24 ].
EXPERIMENTAL VACCINES DELIVERED BY THE MUCOSAL ROUTE
Most vaccines have been developed to improve public health by preventing contagious diseases or by protecting against infections that impose a heavy disease and economic burden on patient subpopulations (e.g., hepatitis B). In contrast, relatively few vaccines are under development for nosocomial infections. P. aeruginosa historically has been an important pathogen in patients with burn wounds, hematologic malignancies, and cystic fibrosis. It remains a significant nosocomial pathogen, particularly in the respiratory tract, and its treatment is complicated by the development of resistance to commonly used antibiotics. Although Pseudomonas vaccines have been developed for parenteral use, a number of vaccines have been examined after mucosal administration. A live, attenuated mutant of P. aeruginosa induced opsonic antibodies after i.n. delivery to mice and protected against fatal pneumonia [25 , 26 ]. Another strategy used oral vaccination with an attenuated Salmonella vector that expressed the Pseudomonas O antigen. This vaccine protected mice against fatal pneumonia when given orally but not parenterally [27 ]. Finally, an enhanced immune response was observed with oral administration of a live, attenuated Salmonella vector expressing the Pseudomonas outer membrane protein (OMP), which was followed by a systemic boost (i.e., prime-boost) [28 ]. Thus, there is precedent for the development of vaccines that target nosocomial infections.
OTHER VACCINES FOR NOSOCOMIAL INFECTION THAT MERIT FURTHER STUDY
For decades, investigators have sought therapeutic strategies to prevent or treat sepsis, a leading cause of death in ICUs. Initially, interest focused on the use of antibodies directed against widely conserved regions in the core region of Gram-negative bacterial LPS. Experimental vaccines were devised from heat-killed bacteria [E. coli O111, Rc chemotype (J5 mutant) or Salmonella minnesota, Re chemotype], in which the core LPS regions were exposed to the immune system (i.e., the LPS lacked the immunodominant O side-chain). An impressive reduction in death from Gram-negative bacterial infections was reported among patients who received the immune serum compared with those receiving preimmune serum [29 ]; however, other investigators were unable to reproduce those results, although none of those subsequent studies duplicated the original study. Further, these latter studies may have failed as a result of the inadequate levels of antibody achieved (see ref. [30 ]). As subsequent studies with mAb to lipid A and cytokine modulation were unimpressive, we developed a subunit vaccine from the LPS of the J5 mutant [31 ]. This vaccine induced protection in animal models of sepsis when administered actively or when vaccine-induced antibodies were given passively [31 , 32 ]. This vaccine was well-tolerated when given to human subjects [33 ].
Preliminary studies using this J5 subunit vaccine in a mouse model of bacterial pneumonia demonstrated the induction of IgG antibodies in serum and bronchoalveolar lavage (BAL) fluid after parenteral or i.n. administration; however, IgA was induced at both sites only with i.n. administration (Table 2 ). When challenged with Klebsiella pneumoniae 2 weeks after final immunization, mice were protected against lethal infection (Fig. 1 ). Thus, mucosal immunization with this vaccine could protect against multiple heterologous species of Gram-negative bacteria causing pneumonia and/or sepsis.
|
View this table: [in a new window] |
Table 2. Comparison of IgG and IgA Antibody Responses of J5dLPS-OMP ± CpG When Given by i.n. or i.p. Routes
|
 View larger version (16K): [in a new window] |
Figure 1. Immunized outbred mice ( ) received 1 µg J5 subunit vaccine + 25 µg CpG i.n., and control mice ( ) received sterile saline solution i.n. on Days 0, 14, and 28. On Day 42, 2 weeks after final vaccination, all mice were challenged intratracheally with K. pneumoniae (O1K2) at 5.8–6.3 x 104 CFU. The Kaplan Meier survival curves from two independent experiments are shown; 12 mice per group; P = 0.0148. In additional experiments, administration of CpG alone resulted in a rate of survival similar to that with saline (data not shown).
|
There have been several attempts at passive immunization of patients for the treatment of C. difficile infection. A parenteral vaccine containing C. difficile toxoids A and B was safe and immunogenic in healthy adults and induced high levels of anti-toxin A IgG [35 , 36 ]. When given to three subjects with recurrent C. difficile diarrhea, this vaccine led to clinical improvement in two individuals [37 ]. Preclinical investigations with C. difficile vaccines in animal models have examined various mucosal, transcutaneous, as well as prime-boost routes of immunization [38 39 40 ]. In the absence of toxin A, C. difficile is avirulent [41 ], but a hypervirulent epidemic strain harbors an additional "binary toxin" that is related to the iota toxin of Clostridium perfringens [3 ]. As the disease is caused by the protein toxin(s), in principle, it should be possible to develop a toxin-based C. difficile vaccine for further evaluation.
EFFECT OF TRAUMA ON IMMUNITY
As nosocomial infections occur in critically ill or traumatized patients, immunization against nosocomial infections would be desirable, particularly as these infections may occur later in the course of the hospitalization. There have been many studies in which the immune response of patients with trauma or burn wounds has been evaluated in vitro [42 43 44 45 46 ]. There is a decreased mitogenic response of PBMC after major injury [42 ], perhaps attributable in part to the induction of a TH2 (IL-4, IL-10) cytokine response [43 ]. Further, with trauma, there is a suppression of B cell differentiation and a shift in the IgM/IgG ratio in PBMCs [44 , 45 ]. When B cell function was examined in vitro, Faist and colleagues [46 ] found supranormal IgA and IgG synthesis but a decreased production of IgM following mitogen stimulation. There are few studies, however, which examine response to immunization in patients who suffer from acute burns or trauma.
In a pilot study of 10 victims of acute blunt and/or penetrating trauma admitted to the Maryland Shock Trauma Center (University of Maryland, Baltimore, MD, USA), Campbell et al. [47
] parenterally administed experimental vaccines against K. pneumoniae (24 valent capsular polysaccharide vaccine) and P. aeruginosa (eight valent oligosaccharide-exotoxin A conjugate vaccine) within 72 h of injury. Nine of the patients had a greater than fourfold response to 
18/24 Klebsiella antigens, and 10 patients responded to greater or equal to seven or eight Pseudomonas antigens [47
]. Robust antibody responses were detected by Day 14, the first day examined. Thus, acute trauma victims can make an antibody response to immunization; however, further studies are needed.
PROPOSED STRATEGY FOR NOSOCOMIAL VACCINES
As acutely traumatized patients can mount a robust response to polysaccharide and conjugate vaccines when given parenterally shortly after trauma, other similarly compromised patients may respond to immunizations. As noted, prototype vaccines for some relevant nosocomial infections are available. Experimental vaccines have been developed for the prevention of sepsis in general and P. aeruginosa in particular, and each type of vaccine has shown encouraging findings in preclinical studies. Moreover, experimental vaccines for fungal infections, such as cryptococcocosis and aspergillosis, have been reported. Therefore, it is conceivable that vaccines could be developed for other nosocomial infections, particularly those that have proven difficult to treat with current antibiotic regimens.
In the absence of mass immunization, some may argue that it may be difficult to identify a population that might benefit from nosocomial vaccines. It may be worthwhile to consider a strategy proposed to improve the immunization rates of patients who might benefit from pneumococcal immunization. As hospitalized patients are more likely to be hospitalized in the future, Fedson and others [1 ] proposed that hospitalized patients should be immunized with a pneumococcal vaccine prior to discharge from the hospital. Studies conducted in the United States and in the United Kingdom have documented that nearly half of influenzal and pneumococcal infection-related hospital admissions and deaths occurred in patients who were hospitalized within the recommended time interval for immunization. In one study, only two of 112 patients admitted for influenza infection were immunized within the preceding 12 months. In the case of pneumococcal pneumonia, it has been estimated that vaccinating 60–100 discharged patients would prevent one hospital readmission [48 ]. Thus, the strategy of immunizing patients to prevent influenza and pneumococcal infection has been referred to as "Sutton’s Law of Prevention." This concept was adapted from the reply of a bank robber, Willie Sutton, who when asked why he robbed banks, replied, "That’s where the money is." In the case of preventing influenza and pneumococcal infection, immunization of hospitalized patients is also "where the money is" [1 ].
A similar rationale could be applied for immunization with vaccines for nosocomial infections were the vaccines available. The same patients who benefit from pneumococcal immunization would be good candidates for receipt of nosocomial vaccines [49 ]. In such a scenario, a prime-boost strategy might have a particular appeal. Upon discharge from the hospital, one could deliver a vaccine to a mucosal site. Given the needle-free administration, these vaccines should be widely acceptable, and one can more easily give multiple vaccines or combinations of vaccines than by the parenteral route. These mucosally administered vaccines, which may require mucosal adjuvants, would prime the immune system. Upon subsequent readmission, one could then administer a parenteral boost (e.g., with a vaccine for C. difficile, S. aureus, pneumonia, or sepsis). These patients would be expected to mount a more rapid immune response that could ameliorate the clinical course of nosocomial infections, especially as these infections often occur later in the course of hospitalization. Studies would be required to examine the effect of different intervals between mucosal prime and systemic boost on the immune response. In this regard, it is instructive that even with prompt initiation of antibiotic therapy, penicillin had no beneficial impact on mortality within the first 5 days of pneumococcal infection [50 ]. This suggests that some enhanced immune response might be desirable, even for infections for which we think there is adequate drug therapy. Such a strategy may also reduce the colonization and in-hospital transmission of nosocomial pathogens, as we have already seen following immunization of children with the pneumococcal conjugate vaccine [9 ]. This approach could also include additional mucosal or systemic vaccine boosts by the primary care physicians in the follow-up care after initial hospitalization.
CONCLUSIONS
The immune response of critically ill patients in general and the mucosal immune response in particular require further exploration. This includes studies to determine the optimal sites of immunization and the delivery systems. It is possible to devise strategies that optimize the immune response to vaccines of nosocomial interest. Given the need and feasibility of immunizing critically ill patients, a more vigorous attempt to develop vaccines against nosocomial infections should be pursued aggressively.
ACKNOWLEDGEMENTS
This work was supported by NIH 2RO1 AI42181-04A1 (A.S.C).
Received June 9, 2007; revised October 4, 2007; accepted October 5, 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
