Pathogenesis of hantavirus infections

INTRODUCTION — The genus Orthohantavirus ("hantavirus") is comprised of a group of more than 30 distinct species of rodent- and insectivore-borne viruses of the family Hantaviridae and more are still being discovered. Two major forms of hantavirus disease are recognized, hemorrhagic fever with renal syndrome (HFRS) and hantavirus cardiopulmonary syndrome (HCPS; also called HPS).

Among the agents of HCPS, the most severe forms are associated with Sin Nombre virus and the southern (prototypical) form of Andes virus; slightly milder forms are caused by the northern form of Andes virus (Andes-Nort), Laguna Negra virus, and Choclo virus. Rio Mamore virus and the closely related Maripa virus are emerging as agents of HCPS in South America [1-5].

The known etiologic agents for the more serious forms of HFRS are Hantaan virus and Dobrava virus; a form of intermediate severity is caused by Seoul virus, and a milder form of HFRS is caused by a vole-borne hantavirus, Puumala virus [6].

The pathology and pathogenesis of hantavirus infections, with a special emphasis on HCPS, will be reviewed here. The composition of these enveloped viruses and their epidemiology, clinical manifestations, and prevention are discussed separately. (See "Epidemiology and diagnosis of hantavirus infections" and "Kidney involvement in hantavirus infections" and "Hantavirus cardiopulmonary syndrome".)

OVERVIEW OF INFECTION — Hantaviruses appear to exhibit similar replicative kinetics and to elicit similar immune responses in both rodents and humans, but rodent infection leads to a persistent state of infection with no overt clinical or pathologic consequence. By contrast, human infection often leads to overt signs and symptoms, subsequent resolution of disease and viremia, and probable immunity to future infection with the same viral strain [7,8]. Levels of plasma viral ribonucleic acid (RNA) appear to be higher in patients with severe clinical manifestations [8]; titers decline during convalescence [9]. The possibility of a persistent state in humans has been raised by a study of immune responses in 78 patients with Andes virus infections in which both T cell and neutralizing antibody responses failed to decline in intensity during years of convalescence [10]. Viremia was not present in convalescent subjects, so the location of any long-term reservoir of Andes virus replication within the host, if present, is not known. Further study is necessary to elucidate whether the persistence of these responses is due to ongoing infection.

Tissue tropism — The hantaviruses show very similar tissue tropism in both human and rodent hosts, with consistent involvement of the vascular endothelium of the heart, kidney, lung, and lymphoid organs [11]. Involvement of the central nervous system is rare.

Neither rodent nor human infection leads to overt cytopathic damage by direct viral involvement, consistent with the pattern observed in most in vitro studies. However, the possibility that hantavirus replication might stimulate apoptosis has emerged in some but not all studies of virus infection in cultured cell lines [12-15]. Thus far, there is no evidence for increased apoptotic activity in human or animal tissues.

Cell entry — Hantaviruses use beta-3 integrins to enter endothelial cells, which leads to dysregulation of endothelial cell migration. Since beta-3 integrins are associated with changes in vascular permeability and maintenance of vascular integrity, hantavirus infection of endothelial cells may play a central role in the disease process of hantavirus cardiopulmonary syndrome (HCPS) [14,16].

Entry of pathogenic hantaviruses into cells appears to require the presence of the beta-3 integrin polypeptide, which in endothelial cells is complexed with alpha-v to make alpha-v/beta-3 heterodimers [17,18]. The entry of pathogenic hantaviruses can be blocked both by vitronectin in a non-RGD (arginine-glycine-aspartic acid)-dependent process, the natural ligand for alpha-v/beta-3, and by anti-beta-3 monoclonal antibodies [19,20]. Investigators have also linked hantavirus entry to the levels of membrane cholesterol in target cells, leading some to propose that drugs that decrease cholesterol levels might prove beneficial in controlling infection [21].

Infected endothelial cells are also rendered dysfunctional by direct toxicities of the virus, such as immune tyrosine activation motifs in the viral Gn glycoprotein, or other virally induced processes that directly inhibit the function of endothelial cells [16,22]. Hantaviruses can enhance the permeability of endothelial cells by sensitizing them to the effects of vascular endothelial growth factor, suggesting a possible in vitro correlate of the observed permeability of vascular endothelial cells during infection in humans [23].

Other molecules have been identified that enable hantavirus entry into cells including DAF/CD55 [24], gC1qR [25], and protocadherin 1 (PCDH1) [26]. The relative importance of these molecules in vivo is not clear, but deletion of the host PCDH1 gene in a hamster model rendered HCPS nonlethal.

EARLY INFECTION — Epidemiologic studies have demonstrated that many hantavirus infections are contracted through the airborne route. After they reach the lung parenchyma, they are taken up in phagocytes and transported to draining lymph nodes. The viruses establish infection in the regional nodes and disseminate to distant organs. The viruses replicate principally in vascular endothelial cells and then establish a secondary viremia [2].

By the time symptoms occur in hantavirus cardiopulmonary syndrome, both viral RNA and antiviral antibodies can be detected in the blood. Patients who succumb quickly to infection display widespread involvement of tissues by viral antigens and RNA [7,9,27-30]. Similar patterns of viral involvement are likely to occur in hemorrhagic fever with renal syndrome, although perhaps involving lower quantities of virus and resulting in lower case-fatality ratios [31].

CARDIOPULMONARY (SHOCK) STAGE — Cardiogenic shock has long been recognized as a key development in the pathogenesis of fatal hantavirus cardiopulmonary syndrome (HCPS) and hemorrhagic fever with renal syndrome (HFRS), with deaths from shock reported from the earliest known outbreaks [32-38]. Hemodynamic measurements in severe cases have demonstrated severe myocardial depression with worsening cardiac indices over time in association with lactic acidosis [33,39]. In patients with HCPS, the cause of hypoxemia is related to increased permeability and is differentiated from cardiac pulmonary edema by its association with low pulmonary artery occlusion pressures and increased protein content of edema fluid [33].

Although it has been recognized that pulmonary edema occurs in HFRS, it has been suspected that such cases can be partially traced to iatrogenic volume overload [36,40-42]. However, it is clear that the agents of HFRS can cause pulmonary edema and that the agents of HCPS can cause renal insufficiency.

INFLAMMATORY MEDIATORS AND THE IMMUNE RESPONSE — Because there is no morphologic evidence for inflammatory involvement of the heart in most cases of hantavirus cardiopulmonary syndrome (HCPS) or hemorrhagic fever with renal syndrome (HFRS), cardiogenic shock must occur through an indirect and strictly functional mechanism, most likely through the action of soluble mediators [43]. Cardiogenic depression in HCPS patients from Brazil has been linked to myocarditis [43]. Such mediators, including the proinflammatory cytokines tumor necrosis factor (TNF)-alpha and interleukin (IL)-1 beta, have long been implicated in septic shock due to bacterial infections [44]. It seems likely that at least part of the myocardial depressant effect of hantavirus infection of the lungs or kidneys is due to the release of soluble inflammatory mediators from the involved organs.

Cytokines — In HCPS, it is suspected that various proinflammatory and antiviral cytokines, such as TNF-alpha, IL-1 beta, and interferon (IFN)-gamma, are agents of a reversible increase in vascular permeability that leads to severe, noncardiogenic pulmonary edema [37,45]. Elevated serum levels of IL-6 and intestinal fatty acid binding protein were independently associated with adverse outcomes in patients with Andes virus in Argentina [45]. One study detected high numbers of cytokine-producing cells in the lungs of patients with HCPS compared with modest increases in patients with non-HCPS acute respiratory distress syndrome, suggesting a role of T cells and local cytokine production [46]. However, studies in an Andes virus HCPS hamster model indicate that the disease occurs even when T cells are depleted [47].

Several proinflammatory and reactive cytokines are upregulated in patients with HFRS due to Hantaan virus (HTNV), including TNF-beta, IL-1 receptor antagonist, IL-6, IL-12p70, IL-10, IFN-gamma inducible protein (IP)-10, IL-17, and IL-2 [48]. A study that compared cytokine responses in HCPS versus HFRS suggested that the former was associated with more activation of innate and proinflammatory responses, especially Th1 responses [49].

Other more subtle findings also argue for the central role of an immunologic process mediated by TNF-alpha and related inflammatory molecules. Both conventional and electron microscopic studies fail to show lytic attack of endothelium by the hantaviruses [50]. Pleural fluid and tracheal aspirates in HCPS resemble plasma in appearance and in the content and composition of protein in these fluids [51]. The observation that, at least for HCPS caused by Sin Nombre virus, there is virtually no loss of red blood cells into the pleural fluid demonstrates that the capillary leak allows only proteins to exit the vascular space, which is not consistent with gross lysis of the endothelial layer.

T cell responses — For both HCPS and HFRS, investigations support a role for CD8+ and CD4+ cytotoxic T cells in the disease process [52-57].

By extension from other acute viral infections, it is possible that many of these T cells are directed against antigenic epitopes in the virus; however, only a very small number of specific antiviral T cell clones have been identified from patients with hantavirus infections [52-54,58].

The T cell response may be protective, at least in HFRS. This was suggested in a study of 18 patients with HFRS in which the combined frequencies of hantavirus-specific T cells five to eight days following fever onset were significantly higher in patients with mild or moderate HFRS compared with those with severe HFRS [59].

Whether T cells are intimately involved in pathogenesis is a subject of controversy. In an animal model of HCPS in which Syrian hamsters are infected with Andes virus and develop a compatible disease, the course of the disease is not affected if T cells are removed from the circulation (using anti-T cell antibodies) before virus inoculation [47]. In contrast, suppression of multiple arms of the immune system (ie, by administering cytoxic agents and corticosteroids) increases susceptibility to hantavirus disease in hamster models [60].

Dendritic cells — Dendritic cells (DCs) play an important role as antigen-presenting cells in viral infections. In vitro studies demonstrate that DCs activated by HTNV upregulate major histocompatibility complex, costimulatory, and adhesion molecules [61]. In addition, infection of DCs with HTNV leads to the release of proinflammatory cytokines (TNF-alpha and alpha-IFN).

Reactive oxygen species — Reactive oxygen and nitrogen species (RONS) are abundant in the lungs of patients with HCPS and in animal disease models of HCPS but absent in the nonpathogenic infection of deer mice [62]. Evidence for increased production of nitric oxide (NO) has been presented for Puumala virus infection in both humans and in a nonhuman primate model [63-65]. Together, these observations seem to put endothelial cells, macrophages, dendritic cells, and their cytokine and chemokine products at the center of the pathogenic process.

Other mediators of tissue injury — Although investigations have concentrated on cytokine abnormalities and have demonstrated that hantavirus infections in vitro can elicit expression of cytokines, the exact nature of the mediators of tissue injury remain to be determined [2]. Other investigations have examined the following [14,16,61,66-76]:

●The role of the kinin system and deposition of immune complexes

●Viral induction of IFN-stimulated genes (ISG)

●Viral depression of endothelial cell migration

●Plasma resistin levels in correlation to hantavirus disease severity

Increasing evidence supports the role of dysregulated innate immune responses in hantavirus pathogenesis. Steady amplification of hantavirus particles in vivo could contribute to ISG responses and/or production of mediators that stimulate vascular leak in endothelial cells. Even replication-incompetent hantavirus particles elicit strong ISG responses for prolonged periods of time [14,69,74,77-79].

IMMUNOLOGIC CORRELATES OF INFECTION — Hantavirus infections elicit both antibody and T cell responses. Specific antiviral immunoglobulin (Ig)M, IgG, and IgA are usually detectable by the time the patient complains of respiratory difficulty in hantavirus cardiopulmonary syndrome (HCPS) [1].

T cell responses have been less well studied. CD8 T cells are present in peripheral blood during the acute phase of illness that recognize Sin Nombre virus (SNV)-specific epitopes on the virus nucleocapsid protein, suggesting a role for the immune response in pulmonary capillary leak syndrome via production of interferon-gamma [2,12].

Host genetic susceptibility may also play a role in disease manifestation [55,80,81]. Individuals with human leukocyte antigen (HLA)-B*3501 may have an increased risk of developing severe HCPS, suggesting that CD8 T cell responses to SNV contribute to pathogenesis [55]. In contrast, the presence of HLA-B27, which has been associated with slower human immunodeficiency virus (HIV) progression, may convey some degree of protection from severe disease caused by Puumala virus and SNV [80,81].

PATHOLOGY — In the case of hemorrhagic fever with renal syndrome (HFRS), capillary leak is manifest in the retroperitoneum; the kidneys are heavy, edematous, and congested, weighing 50 to 100 percent more than expected [82]. Pathologic examination demonstrates perirenal hemorrhage, tubular degeneration, and an assortment of casts; the swollen and hemorrhagic medulla is generally most heavily involved. Inflammatory cells have been identified in the kidneys [83,84].

In hantavirus cardiopulmonary syndrome (HCPS), capillary leak is overwhelmingly centered in the lungs. Autopsy studies demonstrate that the lungs are heavy and edematous, and tracheal and pleural fluid are abundant [85]. Histologic abnormalities include interstitial pneumonitis with mononuclear cell infiltrate, hyaline deposition in the alveoli, and sometimes diffuse alveolar damage [50,85].

Mononuclear inflammatory cell infiltrates can be seen as well in lymphoid organs, including the spleen, lymph nodes, and liver [85,86]. Many of the abnormal mononuclear cells are enlarged activated cells of the lymphoid lineage that are commonly termed "immunoblasts."

In contrast with HFRS, the kidneys are usually uninvolved pathologically in HCPS. There have been reports of HCPS patients with renal failure, but the detection of viral antigens in the kidney may not necessarily indicate the presence of replicating virus in these tissues. (See 'Immunohistochemistry' below and "Kidney involvement in hantavirus infections".)

Patients with severe HCPS and HFRS are also prone to a disseminated intravascular coagulation-like coagulopathy. In a proteomics study, plasminogen activator inhibitor (PAI)-1 and urokinase plasminogen activator (uPA) were upregulated in patients with end-stage HCPS [87]. However, the significance of this upregulation in the pathogenesis of HCPS-associated coagulopathy remains unclear.

Acute myopia and glaucoma are both described as complications of infection with Puumala virus and have apparently not been subject to study in disease from other hantaviruses [88,89].

Immunohistochemistry — Immunohistochemistry studies have shown that hantaviruses have strong tropism for endothelial cells, especially the endothelium of small blood vessels, in both humans and reservoir rodents [11,50,90,91]. The cardinal feature of both HFRS and HCPS is capillary leak [30,37,38,51,92,93]. The localization of viral antigens in the endothelium by immunohistochemistry (IHC) is important because it places the virus in the small vessels of the lung or the kidney, the site of the lesion of clinical significance. Platelet consumption may be linked to the exposure to the blood of normally subcellular layers of matrix and basement membrane that lie beneath the dysfunctional microvascular endothelium.

IHC also shows viral antigens to be disseminated widely throughout the body, including clinically uninvolved tissues such as glomerular endothelium [11,50]. Deposition in renal tissues may occur as a consequence of viremia and does not necessarily imply ongoing viral replication [94]. Hantaviral antigens have also been found within follicular dendritic cells, macrophages, and lymphocytes [50].

In contrast, in HFRS there are more data supporting the presence of replicating virus in the kidneys. It is sometimes possible, for example, to visualize inclusions and virions directly by electron microscopic examination in the renal tubular epithelium of patients with HFRS due to Hantaan virus (HTNV) [95]. Similar involvement of endothelium is observed with the agents of HFRS [83,96].

Less is known about other related hantaviruses. Involvement of macrophages has been observed but is exceedingly rare during infection by the New World North American forms [97]. Pituitary hemorrhage has been observed in necropsy examinations of patients with HTNV and Puumala virus infections [98-100].

SUMMARY

●Microbiology – The genus Orthohantavirus ("hantavirus") is comprised of a group of more than 20 distinct species of rodent-borne viruses of the family Bunyaviridae. Two major forms of hantavirus disease are recognized, hemorrhagic fever with renal syndrome (HFRS) and hantavirus cardiopulmonary syndrome (HCPS; also called HPS). (See 'Introduction' above.)

●Clinical course – Human infection with a hantavirus often leads to overt signs and symptoms, subsequent resolution of disease and viremia, and probable immunity to future infection with the same viral strain. The hantaviruses show similar tissue tropism in both human and rodent hosts, with consistent involvement of the vascular endothelium of the heart, kidney, lung, and lymphoid organs. (See 'Overview of infection' above.)

●Cell entry and relation to enhanced vascular permeability – Hantaviruses use beta-3 integrins to enter endothelial cells, leading to dysregulation of endothelial cell migration. Since beta-3 integrins are associated with changes in vascular permeability and maintenance of vascular integrity, hantavirus infection of endothelial cells may play a central role in the disease process of HCPS. (See 'Cell entry' above.)

●Transmission and pathogenesis – Hantavirus infections are contracted through the airborne route. After they reach the lung parenchyma, they are taken up in phagocytes and transported to draining lymph nodes. The viruses establish infection in the regional nodes and disseminate to distant organs. The viruses replicate principally in vascular endothelial cells and then establish a secondary viremia. (See 'Early infection' above.)

●Cardiogenic shock – Cardiogenic shock has long been recognized as a key development in the pathogenesis of fatal HCPS and HFRS, and death from shock may occur. (See 'Cardiopulmonary (shock) stage' above.)

●T cell immune response and its role in pathogenesis – In hantavirus infections, the immune system, both innate and acquired, is suspected to be a determinant of the magnitude of the capillary leak and its associated morbidity. Infected microvascular endothelial cells may constitute a target of attack by virus-specific T cells, although it is difficult to explain the syndrome as a solely T cell-mediated disease. These antiviral T cells sustain functional damage through viral induction of interferon response genes as well as through direct cytotoxicity. Antiviral T cells collaborate with tissue macrophages to establish a milieu of proinflammatory cytokines as well as chemokines that potentiate direct viral toxicity through the formation of intercellular gaps between endothelial cells. The gaps, in turn, permit the passage of plasma proteins into the interstitium. (See 'Inflammatory mediators and the immune response' above.)

●Other important factors in pathogenesis – Tumor necrosis factor-alpha and interleukin-1 beta are likely to be important factors in the pathogenesis of HCPS, given their abundant expression in HCPS and their known abilities to promote vascular leakage. (See 'Cytokines' above.)