Septic Shock is a serious circulatory disorder often characterised by a whole- body inflammatory state and the systemic response to infection (Titheradge, 1998), with the most common cause being the contamination of blood with bacteria. Septic shock is defined as 'sepsis with hypotension' which develops in almost half of all septic patients as a complication with a mortality rate of 40-60% (Titheradge, 1998).
Septic shock often results in a progressive failure of the circulation to provide blood and oxygen to vital organs of the body resulting in impaired tissue perfusion and oxygen extraction (Thiemermann, 1997). The key symptoms include a severe fall in blood pressure (hypotension) with hypo-reactivity to vasoconstrictor agents (vasoplegia) which may lead to the dysfunction or failure of major organs including lungs, liver, kidneys and brain (multiple organ dysfunction, MODS) and ultimately death (Goligorski et al., 1997).
Presently it is widely assumed that septic shock rarely shows similar symptoms in affected individuals and therefore it is difficult to detect and then consequently treat it (Groeneveld and Thijs, 1986). It is, however, also agreed that most of the therapeutic interventions invariably focus on the primary aim of fighting the refractory hypotension by the use of aggressive fluid infusions, glucocorticoids, large doses of vasoconstrictors (Baumgartner and Calandra, 1999) and occasionally renal replacement therapy (Wheeler & Bernard, 1999). Yet these interventions do not offer consistent success (Parratt, 1997). In recent times, our understanding of the pathophysiology of septic shock has developed significantly through experimental and clinical trials, though the discovery of a suitable treatment with therapeutic efficacy is proving elusive (Baumgartner and Calandra, 1999)'. This is probably because of the heterogeneity of the clinical situations and the differences in host response to identical pathogens. Moreover evidence suggests that different pathogens which cause septic shock respond differently to the conventional treatments. Still, no effort has been made to treat patients according to the nature of the infecting organism (Gao, Anonymous, 1992).
During the early 1990s nitric oxide (NO) emerged as a potentially substantial step towards the treatment of septic shock. This finding directed scientists to carry out numerous clinical trials and animal experiments with the objective of finding out more about the interrelation of NO and Sepsis (Cobb, 1999; Kilbourn, 1999). Later, the discovery of nitric oxide synthase (NOS) made this enzyme the primary target of therapeutic agents (Rosselet et al., 1998). However, recently a substantial amount of literature has been published with evidence contradicting the discoveries of earlier studies. Today more than a decade has elapsed without any resolution to the matter and even in this 21st century Septic Shock is invading the developing countries rapidly (Cobb, 1999; Kilbourn, 1999).
This problem of a lack of significant advances in this field can be highlighted by the fact that 17 years have passed since the final publication of the Consensus Conference on sepsis and sepsis related syndromes (Rangel-Frausto, 2005). This paper conclusively defined sepsis and it's symptoms with the hope of finding a suitable treatment in order to eradicate the disease (Rangel-Frausto, 2005). Today, however, the situation has not changed significantly with up to 750,000 new sepsis cases every year and 215,000 annual deaths in the United States (Trzeciak et al., 2008, Mitchell M. Levy, 2007). Additionally, in the population that survives such attacks, there is considerable morbidity with many scoring low in health related quality of life assessments (Kaarlola et al., 2003, Perl et al., 1995). Hence the '5 million lives' campaign instigated by the US Institute of Healthcare Improvements, aims to minimize the prevalence of nosocomial sepsis (sepsis originating in a hospital) (Gao et al., 2008). This campaign aimed specifically at increasing safety and transforming the quality of hospital care (McCannon et al., 2007). Moreover, the Surviving Sepsis Campaign aims to improve the quality of life of septic patients using the best evidence available currently (Dellinger et al., 2004). Though the world mortality rates have declined in recent times, sepsis is gradually becoming more prevalent in the elderly in the developing countries (Gao et al., 2008).This is mainly due to the extended longevity of patients with chronic illnesses, the increased occurrence of immunosuppression, and the more frequent use of invasive procedures (Bone, 1991; Parrillo, 1993). Sepsis, severe sepsis, septic shock and multiple organ failure still dominate the mass cases of non coronary intensive care units (ICU's).
This essay will attempt to explore the fundamental mechanisms leading to tissue and organ damage in septic shock through the investigation of a case study. Following this, the discovery and general biology of nitric oxide (NO) shall be discussed, and the experimental evidence implicating NO as an effector in sepsis, will be examined in detail. The results obtained from various pharmacologic interventions directed at NO in animal studies will also be considered. This report will also give an account of conventional and innovative treatments for the management of septic shock. Finally, data from the few available relevant clinical trials will be reviewed and possible future avenues of interest will be discussed.
Systemic Inflammatory response syndrome and Septic Shock (Clinical Manifestation)
Sepsis, severe sepsis and septic shock were inaccurately defined until a recently reviewed consensus conference in 1992 (Bone et al., 1992, Dellinger et al., 2004). In this meeting the term ''systemic inflammatory response syndrome'' (SIRS) was invented insinuating a clinical response arising from a nonspecific tissue damaging event (insult) (Rangel-Frausto, 2005).
Sepsis as mentioned earlier is the result of substantial release of inflammatory mediators in response to an infection. However, often the same mediators are released in the absence of a documented infected in several medical conditions such as cardiopulmonary bypass (Wan S. et al., 1997), pancreatitis and trauma (Shanley TP. et al., 2006). Due to this, the early definitions of sepsis or septic shock were found to be misleading hence a North American consensus conference proposed a new terminology. In this conference ''sepsis'' was defined as a condition in which critically ill patients meet criteria for SIRS, in the context of infection. SIRS was defined as two or more of the following: 1) heart rate >90/min; 2) temperature >38°C or <36°C; 3) white blood cell count >12,000 or <4,000 cells mm-3 or the presence of more than 10% immature neutrophils 4) respiratory rate >20/min or a PCO2 <32mmHg (Rangel-Frausto, 2005). Critically ill patients who have a confirmed infection and meet the criteria for SIRS are said to have sepsis. Patients with sepsis can then progress to the development of cardiovascular and other organ failure while also having hypotension (Shanley TP. et al., 2006) and evidence of hyperfusion, are said to have "septic shock" (W. J.Reyes et al., 1999). In recent time it has been established that sepsis can also occur in the absence of a documented infection with the presence of one or more of the following: 1) Cardiac index >3.5 L min-1 M-23; 2) hyperglycaemia (plasma glucose >120 mg/dL) in the absence of diabetes; 3) significant oedema or positive fluid balance (20 mL/kg over 20 h); 4) inflammatory variables: plasma C-reactive protein >2 SD above the normal value or plasma procalcitonin >2 SD above the normal value and 5) mixed venous oxygen saturation (SVO2) >70% . (Dellinger RP et al., 2004). SIRS is generally not considered as a disease and its recognition does not provide any clinical conclusion. However, since it has proven to be a very useful for the identification of sepsis and its sequelae namely severe sepsis and septic shock it is regarded as a very important tool (Shanley TP. et al., 2006).
A 3-year-old boy was admitted to the intensive care unit because of fever, hypotension, and lethargy. A purpuric rash was noted on his arms and legs. Arterial blood gas analysis demonstrated hypoxemia and metabolic acidosis. The arterial lactate level was 10 mmol/L. He was intubated, resuscitated with crystalloid solution, started on broad-spectrum antibiotic therapy, and given dopamine to maintain a MAP above 55 mmHg. His chest radiograph revealed bilateral interstitial-alveolar infiltrates with a left sided predominance. Gram stain of the cerebrospinal fluid showed gram-negative cocci. The dopamine infusion was increased to 18µg kg-1min-1 for persistent hypotension and oliguria. Pulmonary and radial arterial catheter data yielded the following blood pressure, 85/30 mm Hg; MAP, 48 mm Hg; heart rate, 140 beats per minute; CVP, 10 mm Hg; pulmonary artery pressure, 22/14 mm Hg; PCWP, 12 mm Hg; cardiac index, 2.5L min-1m-2; and SVRI, 1226 dyne-sec-cm-5m-2. The boy was treated with additional saline and an epinephrine infusion, which increased the MAP, cardiac index, and urine output. Arterial lactate levels decreased over the next 12 hours. Subsequent cultures of blood and cerebrospinal fluid grew Neisseria meningitidis.
In this case the patient with acute organ failure and hypotension was diagnosed with septic shock. The young boy with meningococcal septic shock had a borderline-low cardiac index. Due to this a decrease in the cardiac index in this patient by the administration of an NOS inhibitor might not be desirable. This is suggestive of the fact that NO or its synthesising means cannot always be targeted in order to battle septic shock. I will attempt to go into further details with regards to this statement as we go along the essay.
Nitric Oxide discovery:
Joseph Priestly first identified gaseous NO in 1772. NO also named Nitrogen Monoxide is a very simple and tiny molecule consisting of one nitrogen and one oxygen molecule. Unknown of its fundamental physiological roles in the mammals, until the 1980's NO was widely considered as a mere toxic atmospheric pollutant (Konstantin J. Ovodov et al., 2000).
In the 1980's researchers were examining how blood vessels expand (dilate) hence regulate the mean arterial blood pressure. Dilation of blood vessels, also termed vasodilation is a very important physiological response which partly regulates the blood pressure. By increasing the diameter of blood vessels, vasodilation causes the blood to travel more freely due to lower resistance (RF Furchgott, 1980). Since the blood vessel lumen widens during vasodilation, the blood imposes less outward pressure on the vessel wall hence reducing the blood pressure. In opposition, vasoconstriction reduces the diameter of the lumen increasing the BP. These physiological responses occur all the time in the human body regulating the BP and therefore are one of the most fundamental mechanisms of the human body.
Dr Robert Furchgott and his group, later in 1980's investigated the role of acetylcholine in the smooth muscle relaxation and found that relaxation only occurred if a special class of cells called endothelial cells were present (RF Furchgott, 1980). These cells line the interior surface of blood vessels, forming an interface between circulating blood in the lumen and the rest of the vessel wall. Behind the endothelial cells are the smooth muscle cells which either relax or contract thus regulating the vascular tone (RF Furchgott, 1991).
The same research group also discovered that smooth muscle were only able to vasodilate the blood vessels in the presence of endothelial cells. This indicated that there was some kind of factor that was being released by the endothelial cells which was involved in the dilation of the blood vessel. This factor was named Endothelium Derived Relaxing Factor (EDRF) and subsequently specified as NO (S Moncada et al., 1997).
In 1977, Ferid Murad independantly investigated the mechanism of action of nitroglycerin and found that it worked by inducing the release of NO which in turn was able to cause relaxation of smooth muscle cells (F Murad et al., 1977).
Louis Ignarro in 1986 finally resolved the whole perplex of EDRF and NO by declaring that EDRF was in fact NO. It was stated that both molecules showed identical properties when he compared gas Nitric Oxide and EDRF (Ignarro, L. J. Et al. 1987). In 1998, Nobel Prize in Physiology and Medicine was awarded to Drs. Robert Furchgott, Louis Ignarro, and Ferid Murad for their discoveries that vascular endothelial cells make nitric oxide (NO) and that such endothelium-derived NO stimulates cyclic guanosine monophosphate (cGMP) synthesis in the underlying vascular smooth muscle, causing relaxation (Kilbourn, 1999).
Since the discovery of NO, vast number of its' physiological roles in normal conditions have been reported including in the immune system, nervous system, reproductive system and other cellular functions. It has also been found to play important roles in variety of species ranging from mammals, to insects and plants.
Role of Nitric Oxide in Biology
Since it was first discovered to play a role in the dilation of blood vessels many new roles for Nitric Oxide (NO) have been discovered. In human body, NO is metabolised by its diffusion into red blood cells where it oxidizes the ferrous iron of oxyhemoglobin yielding methemoglobin and nitrate ions (NO3-) (J.M. Hevel et al., 1994 and Konstantin J. Ovodov and Ronald G. Pearl, 2000). This meachanism limits local NO build up and is particularly important in keeping NO concentration in naomolar range, at least in nonhydrophobic compartments (i.e. outside cell membranes) (Beckman & Koppenol, 1996). Nitric oxide has been found to be produced by effectively every cell type in vivo and plays an important role in both controlling the normal function of cells as well as in regulating larger scale processes such as the reproductive, immune and nervous systems. A few of these biological functions for NO are described in more detail below.
The Immune System
Nitric oxide plays many important roles in the immune system although it is expressed in many cell types following endotoxin or pro-inflammatory cytokine treatment (C. Nathan et al., 1991 and R.G. Knowles et al., 1994). NO, in immune system, is produced in high amounts from specific cells called macrophages. Proceedingly to an infection, chemicals known as cytokines are release in vivo which activate the cells of the immune system such as macrophages, and help guide them to the site of infection (S. Moncada et al., 1991 and C. Nathan et al., 1991). NO produced by the macrophages is toxic to the bacteria and play an important role in their destruction (Fig 1). The production of nitric oxide in this way also helps protect against other types of infection including parasites and viruses (S. Moncada et al., 1991).
However, too much nitric oxide production can not only lead to septic shock but has also been implicated in conditions where the immune system is too active - autoimmune diseases like arthritis (Jang D and Murrell G A, 1998).
The Nervous System
Nitric oxide has been shown to be involved in both the peripheral and central nervous system. The three nitric oxide producing isotypes of enzyme (iNOS, eNOS, nNOS) (see section...) in humans, one isotype (neuronal NOS (nNOS)) is found almost exclusively in the nervous system (Forstermann et al., 1995). nNOS is thought to be involved in promoting the transfer of interneuronal nerve signals. This is thought to take place by the stimulation of exocytosis (release) of endogenous chemicals called neurotransmitters of one neurone (Moncada et al., 1991; Forstermann and Kleinert, 1995). These NT then diffuse across the synaptic cleft (gap between neuronal terminals) and stimulate the neighbouring nerve cell terminal to transmit the signal (Otto Loewi, 1961). NO has shown to play a substantial role in diseases of the nervous system like Alzheimer's and Parkinson's. In both diseases, the inhibition of NO has shown to slow down the progression of the disease in mice (Weill Cornell Team, 2005 and Johns Hopkins et al., 2004).
Nitric oxide is one of the most important molecule in the course of reproduction and is involved in many aspects of it. As well as dilating the blood vessels and thereby helping to regulate maternal blood pressure, NO is also involved in implanting the early embryo in the uterus (Rogers, 1995). During pregnancy, nitric oxide is also suggested promote angiogenesis (a process in which new blood vessels are formed) (RayChaudhury et al., 1996). It is also known to play a role in the survival of trophoblasts (form placenta) (Enders et al., 1978). Furthermore deficiency of NO has been found in patients with preeclampsia (a medical condition in pregnancy) suggesting its partial role in growth of embryo (Yallampalli et al., 1994; Liao et al., 1996). In addition, drugs (Viagra) used to treat erectile dysfunction also affect nitric oxide signalling.
A notable number of cellular activities can be controlled by nitric oxide including cell division, cell movement and cell survival.
The majority of cells in human body have the ability to undergo programmed cellular death. This is a self-destructive mechanism usually called apoptosis which often occurs when a cell is damaged and beyond repair, infected with a virus, or undergoing stressful conditions such as starvation (Kerr et al., 1972). Cells in these conditions go through apoptosis so that they don't hinder the proper functioning of the rest of the tissue. During apoptosis, the structures of the cell break down in an organised manner, forming a packaged cell that is smaller in size so that it can be easily removed by the cells of the immune system (Kerr et al., 1972).
Nitric oxide was first shown to inhibit apoptosis hence promoting cell survival in human B lymphocytes (Mannick, J. B. Et al., 1994). Subsequently similar finding were reported in an ample number of other cells. However, high doses of nitric oxide also have deleterious effects. They can be toxic to many cell types and can lead to septic shock and multiple organ dysfunction syndrome (MODS) in which case NO causes cell death instead of promoting cell survival (Beal & Cerra, 1994).
Nitric Oxide and Inflammation (Pathophysiology of septic shock)
This section will deal with mechanism by which wall fragments of Gram-negative or Gram-positive bacteria and other inflammatory agents induce nitric oxide synthase (iNOS) in cells and tissues.
Exogenous toxins which enter the circulation stimulate the synthesis and release of a number of endogenous cytokines. During a gram-negative infection which can lead to septic shock, lipopolysaccharide (LPS) and endotoxins present on bacterial wall and many other inflammatory agents bind to a co-receptor (CD14) on the surface of specific immune cells like macrophages, resulting in their activation (J.C. Lee et al., 1996). LPS also bind to LPS-binding proteins which are produced by the liver. These proteins facilitate LPS binding to the CD14 co-receptor of the macrophages. The CD14 co-receptor is activated through the binding of LPS to a toll-like molecule (TLR4), (Re F, Strominger J. Et al., 2001) which is responsible for initiating the transmembrane signaling. TLR2 molecules act in the same way as TLR4, (Leppper PM et al., 2002) though; these are activated by gram-positive bacteria, mycobacteria and yeast. Gram-positive bacteria such as Staphylococcus aureus have further additional wall fragments such as peptidoglycan (PepG) and lipoteichoic acid (LTA). Both PepG and LTA have been shown to synergise to produce the characteristic features of septic shock, MODS and ultimately death in rodent models (S.J. De Kimpe et al., 1995 and G.M. Millar et al., 1997). These effects were not observed with either LTA or peptidoglycan alone, although high doses of LTA can cause circulatory failure but not MODS (S.J. De Kimpe et al., 1995). Gram-positive bacteria may also release other enterotoxins and exotoxins, for example toxic shock syndrome toxin 1, which are involved in the pathogenesis of sepsis (reviewed in (R.C. Bone et al., 1994)).
The additional fragments released by Gram-positive bacteria bind to unknown receptors however, like LPS binding, cause the release of proinflammatory cytokines TNF-a, IL-1ß, and IFN-? (Thiemermann, 1997, Titheradge, 1999). These cytokines as well as IL-6 are often produced in response to immune stimulation of macrophages and monocytes hence also in septic shock patients.
IL-1 and TNF each occur in two forms, a and ß. TNF- a and both forms of IL-1 are made by activated monocytes and macrophages, whereas TNF- ß is made by activated T lymphocytes (Review by J. Saklatvala et al., 1996). There are two receptors for both IL-1 and TNF and the two forms of each cytokine interact with the same receptors. IL-1 a and ß interact with the type 1 IL-1 receptor for signal transduction, whereas type II does not appear to transmit any signal and functions as an inhibitor of IL-1 action (J. Saklatvala et al., 1996). The two types of TNF receptor, p55 (type I) and p75 (type II) have different end effects; p75 mediates the proliferative actions of TNF- a while p55 receptor signals the inflammatory response and apoptosis (J. Saklatvala et al., 1996). Complex interactions between these different mediators produce intense pathophysiological modification, which eventually lead to diffuse tissue injury and ultimately sequential system failure (multiple organ dysfunction syndrome), which accounts for the majority of deaths among patients with sepsis, severe sepsis and septic shock (Beal & Cerra, 1994).
IL-1ß and TNF-a have a very short half life compared to IL-6 and therefore IL-6 is a very good indicator of cytokinemia. The initial studies of septic shock, showed a very strong positive correlation between IL-6 levels and fatal outcome (Casey L. Et al., 1993). NO is equally a very short lived molecule with an estimated in vivo half life of only 0.1 seconds hence again it is hard to measure its levels in order to detect the severity of cytokinemia hence the severity of sepsis or septic shock. There are several molecules that contribute to the pro and anti-inflammatory responses in septic shock (Table 4); however I shall only focus on a few due to the limited word allowance.
In response to inflammatory agents in septic shock, the released cytokines (TNF-a, IL-1ß, and IFN- ?), bind to their specific receptors activating a protein kinase called tyrosine kinase leading to both the activation of the nuclear factor-kB (NF-?B) (a transcription factor) and the phosphorylation of intracellular protein (Gao et al., 2008). A precise mechanism by which these cytokines act was proposed by J. Saklatvala et al. in 1996 however this has yet to be confirmed.
Nitric oxide producing cells contain I-?B which is an inhibitor of NF-?B. For the activation of NF-?B, proteolytic cleavage of I-?B from NF-?B is required which forms NF-?B. This biological change allows the activated NF-?B to translocate to the nucleus, where it binds to the promoter region of the iNOS gene inducing transcription. It has also been reported to induce other inflammatory agents, such as cytokines and leukocyte-endothelial adhesion molecules (Janssen-Heininger et al., 2000). Tyrosine kinase present inside the cell acts as a messenger molecule involved in the proteolytic cleavage of I-?B/NF-?B and hence in the activation of NF-?B and iNOS expression (Hecker M, et al., 1996). In septic shock, the translated products of iNOS mRNA subsequently assemble forming the iNOS protein which in turn causes local NO proliferation (Thiemermann, 1997). Fig 2 shows the signal transduction pathway of iNOS expression in response to inflammatory agents.
The physiological role of iNOS is to enhance the formation of NO (due to iNOS activity), which in turn may contribute to either the pathophysiology of septic shock (clinbical symptoms) or the host defence (Reviewed in Gao et al., 2008). Fig 3 shows a simplified schematic of the Anti Inflammatory cascade in the context of septic shock.
Excess NO produced by iNOS has been reported to both induce and inhibit NF-?B (Kalra et al., 2000; Umansky et al., 1998). In year 2000, it was proposed that low levels of NO may induce further NO production while high concentrations do the opposite exhibiting a feedback mechanisms that would oppose the over expression of genes regulated by NF-?B (Janssen-Heininger et al., 2000). In addition NO at high concentrations competes with O2 at the active site of NOS, thus providing a feedback mechanism of its own synthesis (Griscavage et al., 1995; Rengasamy & Johns, 1993).
Biological Synthesis of Nitric Oxide (Nitric Oxide Synthases)
NOS structure and substrates for NO production:
In mammals, NO is exclusively formed from the enzymatic oxidation of one terminal guanidino nitrogen of the amino acid L-arginine. When expressed in moles, this reaction utilizes 1 mol each of arginine and O2, and 1.5 mol of NADPH, yielding 1 mol of NO, 1 mol of L-citrulline and 1.5 mol of NADP (R.G. Knowles et al., 1994). The reaction sequence involves the generation of an Ng-hydroxy-L-arginine intermediate, followed by the oxidation of Ng-hydroxy-L-arginine in presence of molecular oxygen to form L-citrulline and NO (Dennis J. Stuehr et al., 1991 and R.G. Knowles et al., 1994).
The enzymes that accelerate the reaction above are a family of relatively large heme proteins known as NO synthase (NOS) which resemble cytochrome P450 structurally (M. M. Chan et al., 2001 and Francois Feihl, 2001) (The general mechanism of NO production from NOS is illustrated in Fig 4). All members of this family share a similar homodimeric structure, where each monomer consists of a an oxygenase domain and a reductase domain, separated by a short amino acids (30aa) sequence for the attachment of the Ca2+-binding protein calmodulin. In addition to calmodulin attachment, enzymatic activity requires the presence of four cofactors: FAD, flavin mononucleotide (FMN), tetrahydrobiopterin (BH4), and heme (Francois Feihl, 2001). Fig 5 shows the general structure of the NOS enzymes.
Nitric Oxide Synthase isoforms and their locus in the Human Body:
There are three known isoforms of NOS, each the product of a different gene: neuronal NOS (nNOS or NOS-1), inducible NOS (iNOS or NOS-2), and endothelial NOS (eNOS or NOS-3). nNOS and eNOS were first described in rat cerebellum and bovine aortic endothelial cells, respectively, but their tissue distribution is far wider than suggested by their names. eNOS and nNOS are involved in regulating vascular tone (R.G. Knowles et al.,1992 and S. Moncada et al., 1991) and neurotransmission respectively, whereas iNOS is involved in the immune defence although it is expressed in many cell types following endotoxin or pro inflammatory cytokine treatment (R.G. Knowles et al., 1994).
nNOS is typically expressed in skeletal muscle and myenteric plexuses. eNOS is ubiquitous in vascular endothelium, but may also be found in the kidney tubular epithelial cells, placenta (Forstermann et al., 1995), and neurons (Nathan & Xie, 1994). In these tissues, the expression of nNOS and eNOS is constitutive, although it may be regulated (Michel & Feron, 1997). For instance, the levels of transcript for eNOS in vascular endothelial cells is increased by shear stress (Topper et al., 1996; Uematsu et al., 1995) and exercise (Wang et al., 1997), reduced by inflammatory stimuli such as TNF-a (Nathan & Xie, 1994) and variably affected by hypoxia (Le Cras et al., 1998; Toporsian et al., 2000). In the physiological state, the iNOS isoform is only present at a few locations, notably the respiratory epithelium, the gravid uterus (Nathan & Xie, 1994), and perhaps the ileal mucosa (Hoffman et al., 1997). iNOS expression has been demonstrated in numerous cell types including macrophages, neutrophils, vascular endothelial (Hoffmann et al., 1999), smooth muscle cells, endocardium, myocardium, fibroblast, mesangial cells, renal tubular epithelium (Kunz et al., 1994), neurons, hepatocytes, pancreatic islet cells and astrocytes (Nathan & Xie, 1994). iNOS can be induced by a number factors including UV light; cyclic AMP-elevating agents; trauma; ozone and bacterial products described earlier (see section...). On the other hand, many endogenous agents may oppose cytokine induction of iNOS. These include anti-inflammatory cytokines such as IL-10, chemokines such as monocyte chemoattractant protein-1 and growth factors such as tumor growth factor-ß (Forstermann et al., 1995). In all NOS isoforms, calmodulin binding is an absolutely vital for enzymatic activity. In the cases of eNOS and nNOS, this binding necessitates relatively high concentrations of Ca2+ (see Fig 7), in the range of 0.1-1 mM (Forstermann et al., 1995). In contrast, iNOS is able to bind calmodulin virtually independant of Ca2+ (Table 2). Therefore, once iNOS is expressed, NO synthesis may only be limited by the availability of substrates and cofactors (J. Stuehr, 1990). As NO produced from iNOS predominantly depends only on iNOS expression, it lasts much longer than NO formed from the other isoforms of NOS. In addition iNOS produces much higher concentrations of Ca2+ compared to the consecutive forms of NOS (Cobb et al., 1996). The production of NO by eNOS and nNOS, compared to iNOS, can be controlled relatively easily by decreasing or increasing intracellular Ca2+, whereas iNOS can only be controlled through transcription (Cobb et al., 1996). In most cell types iNOS protein levels are either very low or undetectable. However, stimulation of these cells by cytokines or growth factors, can lead to increased transcription of the iNOS gene, with subsequent production of NO. On the other hand, for the prevention of iNOS expression through endotoxins, TGF-ß (Szabo, 1995) and anti inflammatory glucocorticoids can be administered which lower the magnitude of vascular hyporeactivity. Glucocorticoids such as Dexamethasone inhibit iNOS activity by blocking arginine transport and inhibiting tetrahydrobiopterin biosynthesis (A.J.B. Brady et al., 1992 and Thiemermann C et al., 1993).
Regulation of NO production
In the normal as well as in extreme physiological states (e.g. during infection), nitric oxide is considered as one of the most important signalling molecules in vivo. It is however also highly reactive and highly diffusible due to it being a free radical (one unpaired electron) (see fig 8). It is therefore important that there is strict control and regulation of nitric oxide production. The synthesis of NO within cells can be regulated in several ways such as the cellular distribution of NOS, changes in NOS gene expression, enzymatic activation by phosphorylation and the presence of cellular inhibitors NOS activity.
Intracellular distribution of NOS
Nitric oxide is principally regulated through strict control over the location of NO production.
''The NOS isoforms can be targeted to different regions of the cell, where NO will be produced in close contact with its target proteins. The image below shows the distribution of iNOS (shown in green in image 1) and eNOS (shown in red) in a trophoblast cell. The nucleus is shown in blue. Co-localisation between iNOS and eNOS will show up as a yellow colour'' (Phil Dash, University of Reading).
The image shows that eNOS and iNOS are fairly variably distributed inside the cell, with hardly any yellow colour suggesting very little overlap in their cellular distribution (Phil Dash, University of Reading)."Although both iNOS and eNOS produce NO it is likely that their different cellular distribution will lead to NO interacting with different targets and therefore having different effects'' (Phil Dash, University of Reading).
It is very likely that the distribution of NOS isoforms is an important mechanism for regulating when and where NO is produced. Therefore the current research on NOS mainly focuses on how endogenous signals trigger NOS transport and redistribution (Rahul S. Koti et al., 2005).
Activation of NOS activity
NOS enzyme synthesis is principally regulated by changes in intracellular calcium levels. The constitutive isoforms of NOS, (eNOS and nNOS) have shown to proliferate following increases in Ca2+, and therefore calmodulin levels, in the cell (Rameau et al., 2003). Additionally both nNOS and eNOS are known to show further increases in activity following phosphorylation. Currently, at least 5 regulating phosphorylation sites on eNOS are known which are predominately regulated by the protein kinase Akt (Fulton, D et al., 1999). nNOS has both inhibiting and activating phosphorylation sites. iNOS is the least understood molecule as there is still limited understanding about the post-translational regulation of iNOS. A few studies have suggested that iNOS activity may enhance in response to src mediated tyrosine phosphorylation (Orlicek S. et al., 1999), but more abundant studies have shown that iNOS activity is largely regulated at the transcriptional level and through its intracelular distribution.
Endogenous inhibitors of NOS
NO synthesis can also be inhibited by two endogenous inhibitors called asymmetric dimethylarginine (ADMA) and mono methylarginine (L-NMMA) (Rees DD. et al., 1995). These are competitive inhibitors of NOS formed by the post-translational methylation of L-arginine and are released through hydrolysis (Kilbourn, 1997).
ADMA is usually found free in urine and plasma, while the intracellular concentration of ADMA is about 5 times higher than the extracellular concentration (MacAllister RJ et al., 1996). Researchers surprisingly found that when NOS inhibitors were administered into pregnant rats, symptom such as utero-placental blood flow and reduction in placenta size were observed. These features are often found in patients with preeclampsia which can be reversed by the administration of L-arginine, suggesting a distinct NO-mediated effect (Morris NH et al., 1995). In vivo the cellular concentration of L-NMMA and ADMA is controlled by the cytoplasmic enzyme dimethylarginine dimethylaminohydrolase (DDAH) (Reviewed in James Leiper et al., 1999). Inhibition of DDAH activity leads to elevated ADMA in culture and inhibits endothelium-dependent relaxation. These experiments demonstrate that DDAH is basally active and that inhibition of DDAH leads to local accumulation of ADMA which reaches concentration sufficient to inhibit NOS (James Leiper et al., 1999).
How nitric oxide works in vivo (Cyclic GMP)
The most important and well known mechanism, by which the effects of Nitric Oxide are mediated, is through the production of the second messenger cyclic GMP (cGMP) (Louis Ignarro, 1991). Nitric Oxide stimulates the production of cGMP by interacting with the haem group of the enzyme souble guanylate cyclase (sGC) (Fig 8). This interaction allows sGC to convert GTP into cGMP (Denninger, J. W et al., 1999).
Once produced cGMP can have a number of effects in cells, but many of those effects are mediated through the activation of protein kinase G (PKG) (Kilbourn, 1997). Active PKG is ultimately responsible for many of the effects of Nitric Oxide including its effects on blood vessel relaxation (vasodilation). Activation of PKG by cGMP leads to activation of myosin phosphatase which in turn leads to release of calcium from intracellular stores in smooth muscle cells (Nathan & Xie, 1994). This in turn leads to relaxation of the smooth muscle cells. In severe cases when too much nitric oside is produced, this can result into septic shock. During normal physiological state, nitric oxide is originally produced in the neighbouring endothelial cells, with subsequently diffusion into the smooth muscle cells where NO actiavates sGC and cGMP production eventually causing vasodilation (Kilbourn, 1997).
PKG can also activate a number of transcription factors which can lead to changes in gene expression. This in turn may modify the response of the cell to different stimuli (Reviewed in Sausbier M. et al., 2000).
A group of proteins called phosphodiesterases can convert back cGMP to GTP (see fig 10), hence inhibiting further Nitric Oxide signalling. Presently a number of drugs are known to function as phosphodiesterase inhibitors which can restore Nitric Oxide / cGMP signalling. The most popular drug of this class of is known as Viagra (Boolell M et al., 1996).
Nitric Oxide and Septic Shock (Beneficial vs Detrimental effects of NO)
An ample amount of literature produced over the past 20 years has demonstrated that excess production of nitric oxide is involved in the pathogenesis of septic shock. In a healthy individual, NO formation via eNOS, precisely balances sympathetic-mediated vasoconstriction to achieve normal vascular tone and blood pressure (Szabo, 1995). The decreased systemic vascular resistance (vascular resistance) observed in septic shock is due to the excess production of nitric oxide (Szabo, 1995). During recurrent infections, the nitric oxide-induced loss of vascular sensitivity to myocardial depression and catecholamines can contribut to lethal hypotension (Kilbourn, 1997). NO can be produced in response to any process in which proinflammatory cytokines are released including in cytokinemia following cardiopulmonary bypass. In some patients this can account for the vasopressor-resistant vascular dilatation (Kilbourn, 1997).
NO has a very short half life (0.1s) and therefore decomposes spontaneously to chemically stable nitrate and nitrite ions. The first clues to the involvement of nitric oxide in cytokine mediated shock were first observed when endothelial cells exposed to toxins, accumulated nitrite ions (Kilbourn, 1990). Increased levels of serum nitrate/nitrite (NOx) ions in patients with septic shock were first described in 1991 with numerous subsequent studies proving the same (Kilbourn, 1991; Szabo et al., 1995; Neilly I. et al., 1995; Millar & Thiermermann, 1997; Rosselet et al., 1998). Further studies in which IL-2 was given to cancer patients to boost up their immune system, also revealed an increase in urinary and serum NOx (end product of NO metabolism). In addition, tracer studies with arginine labeled at the guanidino position with stable radioactive isotope of nitrogen (15N), revealed high levels of 15N-nitrate in the serum of septic shock patients, demonstrating conclusively that nitric oxide overproduction was present (Reviewed in (Kilbourn, 1997).
These findings and the discovery that nitric oxide is a vasodilator led to the hypothesis that excess nitric oxide causes the hypotension and blood flow abnormalities of septic shock. In order to prove this, the first ever in vivo experiments carried out were on anesthetized dogs administered with endogenous NOS inhibitors (NMA, NAA) and challenged with endotoxin (LPS or IL-1) (Klabunde and Ritger, 1991; Kilbourn, 1992 ). These studies initially reported a greater increase in blood pressure in animals made hypotensive with LPS, IL-2 or TNF than in normotensive controls, suggesting that these agents were specific therapy for sepsis and cytokine-mediated hypotension (Cobb et al., 1996 and Kilbourn et al., 1996). As sepsis induced hypotension had shown to eventually result in multiple organ dysfunction syndrome (MODS), nitric oxide (cause of septic shock) blockage was considered as a potentially successful approach in treating septic shock patients. Further studies in this field showed similar results which prompted researchers to develop NOS inhibitors for clinical use (Cobb JP et al., 1993). In the past few years, a number of potential benefits of NOS inhibitors have been described (Figure 11). These inhibitors have been described to block negative inotropic effects of cytokines in vitro, suggesting that nitric oxide may contribute to sepsis-induced myocardial depression (Finkel MS et al., 1992). Other in vitro studies have demonstrated the ability of high concentrations of nitric oxide to cause cell death in (Nathan et al., 1992), suggesting that this free radical might cause MODS during septic shock and not only as a subsequent complication. Additionally, NO has shown to promote cytokine-mediated injury by increasing TNF production phagocytes. In response to this, NOS inhibitors may exhibit anti-inflammatory activity (Van Dervort et al., 1994).
In spite of growing literature, discovering detrimental effects of NO, a number of studies have also discovered its beneficial properties (Fig 11). Evidence has shown nitric oxide, to dilate blood vessels, scavenge Superoxide molecules (due to free radical) and block platelet and leukocyte adhesion to endothelial cells. These properties suggests that increased production of nitric oxide during septic shock acts to maintain microvascular blood flow and protect the endothelium from oxidative stress and damage (Stamler JS et al., 1994). A number of studies have demonstrated the ability of iNOS to inhibit the effects of septic shock induced endogenous vasoconstrictors and to reduce the activity of eNOS under these conditions (McNaul KL et al., 1993). Additionally, nitric oxide has antimicrobial activity via direct cytotoxic effects on intracellular pathogens (see section...). These lines of evidence clearly show that NO not only causes vasodilation during septic shock, but has numerous other physiological roles too and therefore a complete inhibition NO would not be desirable if a way of doing so is ever found.
Circulatory Failure in Septic Shock
Cardiovascular abnormalities in sepsis include both cardiac dysfunction and altered vascular tone, leading to a typical hemodynamic pattern of high cardiac output, low systemic vascular resistance, and hypotension refractory to vasopressor agents (Brady, 1995; Parrillo, 1993; Rackow & Astiz, 1993). Reduced myocardial performance in patients with septic or endotoxic shock has been frequently proved and established by clinical studies (MacLean LD et al., 1967 and Parker MM at al., 1984 and Ellrodt AG et al.,1985). Early studies also discovered that purified endotoxin caused reversible depression of myocardial function hence reducing systemic vascular resistance (Suffredini AF et al., 1989). The mechanisms by which NO impair the cardiac muscle are unknown however circulating myocardial depressants were proposed as the causative agents of reduced myocardial function (Ellrodt AG et al., 1985. Myocardial dysfunction affects both the right and the left ventricle, and may limit the adaptive changes in cardiac output (Parrillo et al., 1990). The causative agents were subsequently identified as proinflammatory cytokines involved in shock such as TNF-a, IL-1ß, IFN-? and LPS (Thiemermann, 1997, Titheradge, 1999). A plenty of literature published in the 1990's demonstrated the cardiodepressant effects of TNF-a, IL-1ß, IFN-? and LPS alone and in combination. In 1996, Kumar et al. came up with strong evidence that TNF and IL- 1 (TNF-a and IL-1ß more precisely) synergise to suppress the myocardial contractility in patients with septic shock (Kumar et al.,1996). Later studies using both human and rat parenchymal cells have indicated that a mixture of cytokines (TNF-a, IL-1ß, IFN-? and LPS) act synergistically to induce iNOS mRNA and subsequent production of iNOS protein, while LPS alone has no efects (M. Di Silvio et al., 1996). There was substantial evidence that NO caused vasodilation in septic shock but accumulating evidence currently suggest that likewise overproduction of NO can also impair cardiac muscle function. Early studies demonstrated this by exposing hamster cardiomyocytes to proinflammatory cytokines. The results showed a rapid (within 5 minutes) depression of contractility, an effect inhibited by NOS inhibitor L-NMA (Finkel M et al., 1992). The mechanisms of action of these inotropic effects are thought to include rapid stimulation of nitric oxide from constitutive eNOS/nNOS or rapid production of peroxynitrite by means of the combination of superoxide with constitutively produced nitric oxide(Finkel M et al., 1992). Increased iNOS gene expression and activity have been demonstrated in vivo in the hearts of rats given injections of LPS (Luss et al., 1995). Further studies showed a reduction in myocardial function following IL-1 and TNF exposure in ventricular myocytes. Other experiments also showed a reduction in heart rate, an effect prevented by TGF- ß and glucocorticoids (e.g. dexamethasone). The first in vivo experiments carried out on rats in 1992, showed the same effects on myocardial function which were again able to be prevented by NOS inhibitors (e.g. dexamethasone) (Roberta A et al., 1992). Shall I write more about animal studies?
Risk Factors and Causes of Septic Shock
Septic shock is a disorder of the circulation which can develop in any individual, though older people with chronic underlying medical conditions are at greater risk (Balk RA et al, 2000 and Angus DC et al., 2001). Patients with a weak or suppressed immune system, disrupted defense barriers or indwelling catheters are at even greater risk (Balk RA et al, 2000). Immune deficiency in patients can be caused by a variety of disorders or their treatment. Diabetes mellitus, alcoholism, malnutrition and malignancy are the most common conditions associated with increased risk for infection and subsequent septic shock (Balk RA et al, 2000). Cirrhosis has recently been identified as an immune system weakening disease which can cause septic shock (Foreman MG et al., 2003). The data analysed from the National Hospital Discharge Survey database demonstrated increased risk for sepsis and septic shock in patients with cirrhosis (Foreman MG et al., 2003). Recent studies have shown that patients, who are genetically more susceptible to infections, are at greater risk of the developing septic shock. These studies may explain the heterogeneity observed in evaluation of new therapeutic molecules in patients with septic shock (Holmes CL et al., 2003 and Lorenz E et al., 2002). Every individual possesses a unique genetic makeup which can as a result cause heterogeneity in cytokine expression and synthesis. Genetic polymorphisms have been identified for all the genes (TNF, IL-6, IL-1, IL-10, LPS, CD14, TLR2 and TLR4) involved in septic shock (Holmes CL et al., 2003). This is evident in inviduals, who express high quantities of IL-10 (compared to low IL-10 producing individuals) in response to infection, making them more susceptible to septic shock. ''It has also been observed that in male patients with trauma there is an increased tendency for bacteremia and higher mortality, and a greater likelihood of sepsis development, which is thought to be related to androgen-induced immunodepression and enhanced release of TH-2 lymphokines. In contrast, in female patients there appears to be immunoprotection and an immune-enhancing effect on B cells and macrophages, related to the release of increased TH-1 lymphokines'' (Angele MK et al, 2000).
Micro and Macro-circulation in Septic Shock
The microcirculation is an integrated functional system that helps ensure that tissue oxygen demands meet the delivery throughout the body. Septic shock is often associated with complete failure of this system.
In the early studies, harmful properties of NO became evident from its consequences on the macrocirculation (arterial hypotension). The possible beneficial effects on the microcirculation were completely disregarded at the time. Numerous studies involving iNOS upregulation have repeatedly shown that NO induces relaxation of microvascular tone and hence causes arterial hypotension.71-74 Henceforth, researchers' initially aimed at inhibiting NO production through NOS inhibitors. However, although NOS inhibition was shown to clearly raise arterial pressure in septic shock, (Hollenberg SM, 1993 and 1999) other studies showed, it can simultaneously worsen the microcirculatory perfusion and oxygen transport to tissues (see fig 13) (Shultz PJ et al, 1992; Avontuur JA et al., 1995).
NO has shown to play multifaceted roles in the regulation of homeostasis in the Microcirculation during septic shock. In health, NO maintains microcirculatory homeostasis by regulating microvascular tone, platelet aggregation, leukocyte adhesion, platelet aggregation, microvascular permeability and microthrombi formation (Cambien B et al., 2003; Shultz PG et al., 1992). These features become disrupted in septic shock leading to defects in myocardial blood flow and tissue oxygenation. Recent NOS gene knockout results showed that the iNOS deficient mice reduced the in vivo reactivity of the microcirculation hence demonstration the importance of NO in the microcirculation (Hollenberg et al., 2000). During sepsis and septic shock, NO molecule becomes absolutely crucial to maintaining microcirculatory integrity and function. The proinflammatory cytokines initiate a sharp increase in systemic NO production (Cunha FQ et al., 1994), though the upregulation of inducible nitric oxide synthase (iNOS) is heterogeneously expressed between and within organ systems (Morin MJ et al., 1998 and Cunha FQ et al., 1994). This can possibly occur due to reactive oxygen species which can scavenge NO molecules hence, giving the potential for localized areas of relative NO deficiency in microvascular beds despite an overall excess of NO in the body (Reviewed in Ince C et al., 2005). This has been shown in both human and experimental models of septic shock suggesting the rational behind the heterogeneity of tissue perfusion (Trzeciak S et al., 2007). This may also help explain the diversion of blood flow away from distressed microvessels (microcirculatory shunting) (Fig 13) via opening of arteriovenous shunts within capillary beds (Ince C, et al., 1999). These effects have also been observed in rat models of cegal ligation and puncture. A recent study, however was unable to document any indication of improved organ perfusion when the mean blood pressure of septic patients was raised from 65 to 85 mm Hg with NE infusion (LeDoux et al., 2000).
With the use of intravital video microscopy, researchers have demonstrated that septic shock is characterized by ''decreased microcirculatory flow velocity, an abundance of stopped-flow microvessels, increased heterogeneity of microcirculatory flow, and low density of perfused capillaries'' (Fries M et al., 2006; Trzeciak S et al., 2007). Recently De Backer et al used orthogonal polarization spectral imaging in two studies of the effects of acetylcholine, human activated protein C and dobutamine on the microcirculation in septic patients. All agents were found to increase capillary perfusion, suggestin that the endothelium was still NO responsive (De Backer et al., 2004). A very recent study on mice models showed beneficial effects of inhaled NO on the microcirculatory perfusion. Higher levels of NO were found in the mice following NO inhalation (Nagasaka Y et al., 2008). This information has led to a randomized clinical trial currently underway which is testing efficacy of inhaled NO (Trzeciak et al., 2008).
The majority of septic shock related detrimental effects of the microcirculation have been shown to occur in the microcirculatory unit. This unit is formed of arteriole, capillary bed and the postcapillary venule, where during septic shock, loss of vasomotor reactivity, endothelial cell (EC) injury, microthrombus formation, and disordered leukocyte trafficking can take place (Fig 12) (Trzeciak S et al., 2007).
Theoretical model of capillary perfusion
The figures below illustrate that sepsis-induced microcirculatory dysfunction plays a key role in impairing oxygen transport to the tissues and contributes to tissue hypoxia. '' (A) Healthy state: A cylinder represents the area of tissue that is supplied with oxygen by an individual capillary. The diffusion distance for oxygen in the tissues is shown (small arrow). (B) Sepsis: Intrinsic microcirculatory dysfunction results in nonperfused capillaries (dotted line vessels). This decreases the density of perfused vessels, increasing the diffusion distance for oxygen in the tissues (large arrow)'' (Trzeciak S et al., 2007).
Early Diagnostic Strategies
For the effective treatment of septic shock, its early diagnosis has always been a goal (Bone et al., 1992 and 1997). It was hoped, that through this, an earlier administration of effective therapy would result in improved survival. Traditionally, patients have to wait for the blood test results before any treatment is given. In the mean time several other complications can result making subsequent treatment more difficult. The clinical criteria proposed in the SIRS definition is widely considered as too sensitive. Recent studies have shown that the analysis of clinical laboratory assessments, such as procalcitonin levels, biphasic A2 waveforms, C-reactive protein (CRP), endotoxin, or cytokine levels, as potential early indicators of sepsis (Casey LC et al., 1993; Lobo SMA et al., 2003). Some European centers have used procalcitonin and CRP concentrations for early diagnosis of severe sepsis (Luzzani A et al., 2003 and Lobo SMA et al., 2003). Reports of a number of centres using CRP and procalcitonin levels for early diagnosis of severe sepsis/septic shock have shown strong positive correlations between these molecules and the presence of sepsis (Luzzani A et al., 2003 and Lobo SMA et al., 2003). Several studies have demonstrated that procalcitonin levels can also be used to indicate severity of illness and predicted survival (Luzzani A et al., 2003). Most studies found procalcitonin concentration to better predict sepsis compared with CRP concentration (Luzzani A et al., 2003). However several investigators raise the issue about the inability of these tests to differentiate infectious causes of a systemic inflammatory response (sepsis) from other causes. Low density lipoproteins and Ca2+ associated coupling of CRP produces a biphasic waveform ''for determination of activated partial thromboplastin time with the MD80 analyzer'' (Toh CH et al., 2003). Studies are currently looking into whether this abnormal waveform can serve as a marker for early diagnosis of septic shock. Earlier efforts observed cytokine and endotoxin concentrations, alone and in combination, in an attempt to detect sepsis or septic shock (Casey LC et al., 1993). These studies however only demonstrated the presence of bacterial toxins or cytokines with little correlation between endotoxemia and the underlying pathogen or predict outcome (Marshall JC et al., 2002). Endotoxemia often arises during abnormalities in intestinal barrier function which causes translocation of endotoxin bacteria from the terminal ileum (Balk RA et al., 2000).
Treatment (Objectives in the initial resuscitation of patients with septic shock)
12 years have passed since the surviving sepsis campaign, conclusively specified the following course of treatment for sepsis or septic shock. The most important aim in the managment of septic shock was to identify and control the source of infection via antimicrobial therapy followed up by hemodynamic support, organ support, sedation or analgesia as needed and adequate nutrition (Rangel-Frausto, 2005). Several other treatment strategies have been proposed over the past few years; however these are still under discussion (Dellinger RP et al., 2004). The fundamental goal in the treatment of septic shock is to recover cardiac preload/afterload and contractility in order to improve oxygen delievery to the tissues. In order to do this, early goal therapies (treatment in first 6 hours) have been designed. ''These guidelines/goals include 1) central venous pressure 8 and12 mmHg; 2) mean arterial pressure above 65 mmHg; 3) urine output above 0.5 mL/kg/h; 4) central venous (superior venacava) (SvO2) saturation above 70% (Dellinger RP et al., 2004). A few animal studies have reported strong correlation between these therapies and mortality. These results ofcourse promote the use of these therapies however since clinical trials are still under way, their clinical use is currently prohibited (Rhodes A et al., 2004; Freid MA et al., 2004).
Antibiotics (Conventional Approach)
Over the past 20 years, a great number empiric antibiotic therapy and other clinical trials conducted on septic shock patients were non-specific in terms of the infecting organisms. McCabe et al first reported a reduction in mortality (from 48% to 22%) when appropriate therapy was given (McCabe WR et al., 1962). Recent in vitro studies have shown that there are mechanistic differences between, fungal, viral and bacterial sepsis, and imply that pathogenetic differences may exist between subclasses such as Gram-positive and Gram-negative infections (Gao et al., 2008). These differences can lead to variations in cytokine levels and mortality rates associated with Gram-negative and Gram-positve sepsitic shock. In a bid to prove this hypothesis, a randomized, double-blind, placebo-controlled trial was conducted, in which patients and controls were given soluble fusion protein of TNF-a receptor (Fisher CJ Jr et al., 1996; Cohen J et al., 1996). The results showed no harmful events in patients with Gram-negative infection, whereas patients with Gram-positive infection tended to have increased mortality (Fisher CJ Jr et al., 1996). In contrast, a murine monoclonal antibody directed against human TNF-a tended to reduce mortality in Gram-positive infection, while Gram-negative infections showed the opposite effects (Cohen J et al., 1996). Moreover, Gram-positive infected patients have potentially been harmed in trials of anti-LPS and IL-1 receptor antagonists, while Gram-negative infected patients didn't demonstrate as extreme results (Fisher CJ Jr et al., 1994). It is often assumed that combinational therapy is more effective than monotherapy in treating septic shock, however recent studies have shown that there is no overall increase in mortality (compared to monotherapy) when drugs are given in combination. The antibiotics that are presently used to treat septic shock include Carbapenem, third or fourth generation cephalosporins, or ureidopenicillins with b-lactamase inhibitors (Dupont H et al., 2000).
Haemodynamic Support; Fluid Resuscitation (Conventional Approach)
Septic Shock, as described earlier is caused by extreme vasodilation with a decrease in systemic vascular resistance (Titheradge, 1998). The mechanisms surrounding this vasodilatory shock include increased synthesis of the endogenous vasodilator nitric oxide (NO) as a result of increased activity of iNOS, and a deficiency of vasopressin (Landry DW et al., 2001). There is also an increase in capillary permeability; causing low perfusion pressure and an overall low vascular fluid volume. In some patients this can be exacerbated by diarrhea, vomiting, or other causes of fluid loss (Cook D et al., 2001).
At this point, adequate volume resuscitation can restore normal perfusion and mean arterial pressure. There are still contradictions with regards to the selection of fluid. Some researchers suggest superiority of colloids to crystalloids or albumin while others demonstrate vice versa. A recent randomized, double blind clinical trial showed no difference in mortality in patients treated with saline or albumin solution. The use of albumin however proved very beneficial (Haynes GR et al., 2003).
Corticosteroids (Innovative Approach)
Clinical and animal models of septic shock have shown that high levels of proinflammatory cytokines (produced in response to infectious agents) induce SIRS (Ottosson J et al., 1982; Demling RH et al., 1981). A number of potential beneficial roles of corticosteroid agents in the treatment of severe sepsis and septic shock have been proposed (Table 3) (Balk RA et al., 2000; Angus DC et al., 2001). Animal studies have reported reduced mortality in response to high doses of corticosteroid agents (Hollenbach SJ et al., 1986).
In addition, clinical investigations reported that patients with septic shock with cortisol baseline levels above 34 mg/dL in whom corticotrophin stimulated increase in cortisol was below 9 mg/dL showed an average of 74% survival rate while in the opposite conditions, there was only an 18% increase in survival rate (Ananne D, 2002). This information led to the discovery that patients in that category receiving low-dose hydrocortisone had lower mortality. Prevalence of relative adrenal insufficiency goes from 50% to 75% in septic shock patients. Low doses (200-300 mg/day for 7 days, in three or four divided doses) of hydrocortisone are recommended in patients with septic shock who despite adequate resuscitation with fluids require vasopressors (Ananne D, 2002). In these studies, however, 1-year mortality was not different in patients who received steroids compared from those who only received fluid resuscitation and vasopressors (Ananne D, 2002).
Vasopressor use in Septic Shock
Vasopressors are a group of drugs used for resuscitation of seriously ill patients with critical hypotension. The vasopressors used in septic shock are catecholamines which include the drugs Epinephrine, Norepinephrine, Dopamine, Dobutamine and Isoproterenol (Marc Leone et al., 2008).
Early randomized clinical trials which looked into the effects of vasopressors on septic shock compared the ability of norepinephrine and dopamine to reverse the common abnormalities (i.e. hypotension, organ dysfunction etc.) associated with the disease (Martin C, et al., 1993). The results of these studies showed that, norepinephrine was more reliable and effective than dopamine to reverse the abnormalities of septic shock. Norepinephrine was able to increase oxygen uptake and mean perfusing pressure without causing any adverse effects on peripheral blood flow or on renal blood flow in majority of the patients. In addition, a non randomized clinical trial, found reduced mortality patients treated with norepinephrine than in those treated with dopamine or epinephrine (Martin C, et al., 1993).
In other clinical trials, septic shock patients treated with dopamine showed reduced survival rates (Levy B et al., 2005; Sakr Y et al., 2006). The reports of an observational study including 1058 patients treated with dopamine, showed an increase of 42.9% in hospital mortality rates (Sakr Y et al., 2006).
Other studies demonstrated that dopamine and norepinephrine have similar hemodynamic effects, while epinephrine can damage the splanchnic circulation as a side effect in severe septic shock. This is in agreement with prior studies showing that epinephrine is as effective as norepinephrine or dobutamine, but impairs the splanchnic circulation (Martin C, et al., 1999). De Backer et al later compared the effects of epinephrine, norepinephrine and dopamine on the splanchnic circulation in patients with septic shock (De Backer D et al., 2003). Dopamine was again found to cause detrimental effects which led to its withdrawal from the treatment. Subsequently to this, norepinephrine and then epinephrine were given to either maintain constant mean arterial pressure (moderate shock) or to increase mean arterial pressure above 65mmHg in case of severe shock (De Backer D et al., 2003).
The most recent randomized clinical trial compared the safety and efficacy of dobutamine+norepinephrine with epinephrine alone in 330 patients with septic shock. The results showed no significant difference in mortality rates between the two groups (Annane D et al., 2007). Epinephrine treated patients, however demonstrated a quicker recovery of the circulation, but had significant transient metabolic side effects compared to norepinephrine. There was a trend in favor of epinephrine in the subpopulation of patients with sepsis (Myburgh J et al., 2007). The latest studies, in contradiction to earlier studies, show no difference in the outcome of the disease with the use of different catecholamines suggest further clinical trials need to be carried out with an even larger scale patients (Annane D et al., 2007; Myburgh J et al., 2007).
Timing of NOS inhibition and survival rate
Recent animal studies have shown that mice with overexpressed eNOS are resistant to the deleterious effects of endotoxin (Yamashita et al., 2000) and that the eNOS acts as iNOS in producing venodilation in humans during endotoxemia (Bhagat et al., 1999). These findings highlight the potential hazards of focusing on overproduction of NO as the single most important target when addressing the underlying mechanisms of the rather high mortality rate of septic shock.
Since Septic shock is often described as a disease organised in distinct phases with diverse features, the following study was aimed at finding the significance of timing of NO synthase inhibition.
In this experiment animals were given L-NAME (a non-specific NOS inhibitor) or L-canavanine (a specific iNOS inhibitor), twice at 2 and 6 h after endotoxin injection. The results plotted on survival curves were closely parallel to that of controls (saline) hence suggesting no significant difference from the controls (Fig 15) (Alper B et al., 2001). However, when the mice were treated with another specific iNOS inhibitor, aminoguanidine, twice at 2 and 6 h after endotoxin injection, the survival rate at 24 h was 70%. This was significantly different from the corresponding control value (Fig 15) (Alper B et al., 2001).
Animal Models of Shock (Non selective NOS inhibitors)
NOS inhibitors to treat human septic shock are based on the following three assumptions: (1) NOS inhibitors are safe (2) the hypotension of septic shock is due to excess production of nitric oxide and (3) this excess production of nitric oxide contributes to mortality.
The experimental studies of animals have been instrumental in the understanding of sepsis. In 1990, Kilbourn et al showed significant hypotension in anesthetized dogs challenged with LPS that progressively worsened following 2 hours after LPS was given through bolus. This effect was able to reverse after 20 mg/kg of L-NMMA (non selective NOS inhibitor) was administered. Later when L-arginine was given to the dogs, the effects of L-NMMA (N omega-amino-L-arginine) started to abolish. Moreover the study found that, the hypotensive effects were not observed when nitroglycerine instead of LPS was given (Kilbourn et al., 1990). These initial observations were followed by further studies demonstrating that administration of L-NMMA following IL-1 mediated shock in anesthetized dogs leads to restoration of blood pressure, however with the consequence of lower cardiac output. Nava E et al subsequently reported higher survival rate in mice receiving intermediate doses (30 mg/kg) of L-NMMA (Nava E et al., 1992). Contrary to this, in 1994 it was reported that large doses (300 mg/kg) of L-NMMA increased mortality in anesthetized (antibiotic treated) endotoxemic rats (Evans T et al., 1994). In an earlier similar study, opposite (deleterious) effects were reported (Teale DM et al., 1992). This may be because in the latter study, awake rats challenged with E. Coli were used rather than anesthetized rats challenged with LPS. In another investigation, the effects of another non selective NOS inhibitor, L-NAME (NG-nitro-L-arginine methyl ester) were observed (Fukatsu K et al., 1995). The results showed increased in vivo bacterial growth and higher TNF (an inflammatory mediator) levels. In addition, a further study showed that the detrimental effects of E. Coli mediated shock which, were able to reverse in response to L-NMMA in a conscious mouse. Since the animals used in this study were awake during the whole experiment and exhibited many of the characteristics of human, this was the first study that could be compared to human (Rees DD et al., 1998).
Animal Models of Shock (Selective NOS inhibitors)
In septic conditions, nitric oxide synthase inhibitors might be of therapeutic value, but detrimental side effects (e.g. low cardiac output) have been reported with their use. Therefore, the use of selective inhibitors of inducible nitric oxide synthase might be more suitable. In 1994, Szabo et al, reported that the selective iNOS inhibitor, S-methylisothiourea sulphate, significantly improved renal and hepatic function in rats and increase survival rates in mice (Szabo et al., 1994). Another study which used the selective NOS inhibitor, L-Canavanin showed a reduction in the severity of all the detrimental effects of LPS such as low BP, low cardiac index and high nitrate levels. In addition it increased survival in mice which were given lethal dose of LPS. In contrast to L-canavanine, nitro-L-arginine methyl ester (L-NAME (non selective NOS inhibitor)) increased blood pressure at the expense of a severe fall in cardiac index, while largely enhancing lactic acidosis. This agent did not improve survival of endotoxaemic mice (Liaudet L et al., 1996).
In contrast to nonselective NOS inhibitors, which increase vascular reactivity and arterial blood pressure at all stages of endotoxic shock, the effects of selective iNOS inhibitors were restricted to the delayed, iNOS-dependent stage. When administered to normal rats, selective compounds did not elicit significant pressor responses, confirming a lack of eNOS inhibition, at least when using low doses (Liaudet L et al, 1998). However, the selectivity is only relative, and these compounds may block all NOS isoforms when used at sufficiently high doses.
Although the number of human clinical investigations is understandabale low compared to animal data, they have indicated that the beneficial and detrimental effects of NOS inhibition are very similar to animals. Most of the available clinical investigations were essentially limited to septic patients with severe hypotension, which was not able to reverse in response to vasopressors and fluid resuscitation. Till now, only non selective NOS inhibitors have been administered in human. First time in 1991, Petros et al studied the effects of L-NMMA and L-NAME on two septic patients, followed in one case by continuous i.v. administration of the latter agent (Petros et al., 1991). The results showed that patients became haemodynamically stable within 48 hours and required less vasopressor use to maintain the BP.
Another study conducted by the same research group, 3 years leater, found similar results. The main difference between these trials was that in the former study, patients were administered with 0.3 mg/kg of L-NMMA which resulted in a 7 mmHg rise in BP while in the latter study the same drug was given at a concentration of 1 mg/kg which showed a 22 mmHg rise in BP. In both trials, like in previous animal studies, L-NMMA also caused lower cardiac output (Petros et al., 1994).
A further clinical trial in which septic patients received a fixed dose rate of L-NMMA reported an increase in systemic vascular resistance, associated with a decrease in cardiac index, within the first hour of therapy (Hussein et al., 1999). Subsequently, the systemic vascular resistance returned toward pretreatment values upon reduction of the concomitant norepinephrine therapy. There was a reduction in the levels of plasma nitrate over time during and after the treatment phase (Hussein et al., 1999).
A very recent multicenter, randomized, double-blind, placebo-controlled trial of patients with septic shock treated with the nonspecific NOS inhibitor NG-methyl-L-arginine (L-NMMA), revealed no improvement in mortality of patients with septic shock, although hypotension was improved (Lopez et al., 2004). This was a surprising finding, since most of the beneficial and detrimental effects of L-NMMA treated mice were also observed in human. L-NMMA increased survival rate in mice given lethal dose of LPS.
The reason why this clinical trial failed is unknown, however concerns about patient selection, dose, titration to maintain blood pressure, as well as the nonspecific nature of the NOS inhibitor used have arisen.
Studies in iNOS knockout mice
The main reason for NO overproduction in septic shock is the presence of iNOS in various tissues. This early finding led to further studies on iNOS deficient animals conducted over the past 15 years with the hope to enhance our knowledge about the pathological role of NO. In 1995 three separate groups independently disrupted the iNOS gene by attacking it at different regions. In all cases, expression of iNOS was inhibited and could not be induced under conditions that induce iNOS in wild-type animals (MacMicking et al., 1995; Laubach et al., 1995; Wei et al., 1995). In these experiments, the peritoneal macrophages of iNOS deficient mice produced no NOx and failed to express both the iNOS mRNA (gene) and the iNOS protein. Inducible NOS deficient mice were found to be more sensitive to the intracellular parasite Leishmania major (Wei et al., 1995) and to Listeria monocytogenes (MacMicking et al., 1995) than were the wild-type mice. In the physiological state, these pathogens trigger off the immune response which demonstrates that iNOS is very important in the host defence system. These studies also showed blunted hypotensive response to septic shock and LPS suggesting that NO generated by iNOS contributes to the inappropriate vasodilation and hypotension seen in septic shock (MacMicking et al., 1995; Laubach et al., 1995; Wei et al., 1995). Further studies on iNOS deficient mice have reported reduced in vivo reactivity of the microcirculation. This shows that the constitutive form of NOS (i.e. eNOS and nNOS) can not compensate for the NO produced from iNOS in response to infection during septic shock (Hollenberg et al., 2000). A very recent study revealed that septic iNOS-deficient mice ''exhibited less microvascular leakage than wild-type septic mice despite equivalent increases in leukocyte adhesion'' (Hollenberg et al., 2007). This suggests an important role for nitric oxide in modulating vascular permeability during septic shock.
Studies in nNOS and eNOS knockout mice
Studies on nNOS deficient mice have reported significantly low NO production in the brain. NO levels were measured by spin trapping NO, cGMP levels and NOS enzymatic assays (Irikura K et al., 1995). Neuronal NOS knockout mice demonstrated enlargement of the stomach several times to the normal size, indicating a role of NO produced from nNOS in muscle relaxation of the pyloric sphincter (Irikura K et al., 1995). These mice were also found to be resistant to cerebral ischaemia consistent with a role for nNOS-derived NO in cellular injury following ischaemia (Panahian N et al., 1996).
Endothelial knockout mice show mild abnormalities in vascular relaxation, cardiac contractility and blood pressure regulation and therefore are very useful animal model for endothelial dysfunction. These animals showed higher susceptibility to form neointima in response to vessel injury (Zhang L et al., 1999).
It is evident that excessive production of NO following the induction of iNOS as a result of an increased circulating cytokine concentration makes a major contribution to the development of the characteristic symptoms of septic shock. Nitric oxide is a short-lived diatomic radical formed from the amino acid L-arginine by a group of enzymes called nitric oxide synthases. It activates soluble guanylyl cyclase, which decreases intracellular calcium and facilitates smooth muscle relaxation hence causing vasodilation and therefore extremely low BP in septic shock.
The NO molecule is vital to microcirculatory homeostasis, but various studies have shown that during septic shock when there is an overall excess of NO, the microcirculation lack NO.
Much of our knowledge concerning the involvement of NO in septic shock and the use of non-specific inhibitors of NO synthases has been derived from rodent isolated tissues and rodent models.
More than 10 years have elapsed since nonselective NOS inhibitors were shown to reverse the hypotension of septic shock. The side effects of these inhibitors e.g. inhibition of both eNOS and iNOS) led to the discovery of selective iNOS inhibitors which showed positive results in septic rodents. Clinical trials employing selective iNOS inhibitors should allow us to better understand the role of excessive NO production in human during shock.
NOS knockout mice studies have greatly added to our understanding of the roles of all the isoform of NOS. Inducible NOS deficiency has been reported to suppress the immune system and cause the abnormal vasodilation during shock.
Despite these discoveries and therapeutic approaches, the mortality associated with severe sepsis and septic shock remains unacceptably high. Even after recovery from severe sepsis and septic shock, mortality is higher in patients during the first year of follow-up, as compared with matched control subjects. Current research interests have targeted early diagnosis as a means to improve both short-term and long-term outcomes.