Immunological mechanisms of pentoxifylline in chronic heart failure

Abstract

The progressive syndrome of chronic heart failure (CHF) represents a common disease pathway that may be derived from a host of varying insults (including myocardial ischaemia and infarction, hypertension, viral infection, pregnancy, etc). Despite this multifarious aetiology, a common phenomena observed in CHF patients is elevated levels of tumour necrosis factor (TNF)-α. This has led to the widespread concept that TNF-α is directly involved in the pathophysiology of CHF and as such, attempts have been made to inhibit TNF-α production in this cohort. However, to date, there have been no clear beneficial effects from TNF-α inhibition and indeed trials of direct anti-TNF therapy have provoked worsening of clinical outcomes. Conversely, a possible exception is pentoxifylline (PTX), a putative TNF-α inhibitor with possible (but ill-defined) vasodilatory properties. Several small clinical trials assessing the use of PTX in CHF have suggested beneficial effects on multiple surrogate clinical markers. Interestingly, these trials failed to show a concordant effect on circulating TNF despite the clinical improvement, suggesting other key beneficial properties of this novel agent. This review article provides an insight into the potential beneficial mode of the action of PTX in CHF and calls for more investigation of this interesting agent.

Introduction

Patients with chronic heart failure (CHF) have typically endured an initial insult to the myocardium, which is then followed by a gradual deterioration of cardiac pump function. Those who develop severe CHF, defined as New York Heart Association (NYHA) class IV, carry a particularly poor prognosis.¹ Over the past 20 years, significant advances have been made in the treatment of this condition. Yet, despite current optimal therapy, cardiac output commonly deteriorates over time in a large percentage of CHF patients. This suggests our understanding of the pathophysiological mechanisms involved in CHF has yet to be fully elucidated. A potential process involved in CHF is immune activation. This has been proposed following a series of observational studies that identified heightened activity of the immune system in patients with CHF. In particular, CHF patients appear to have elevated plasma levels of pro-inflammatory cytokines including tumour necrosis factor-α (TNF-α).² Levels of TNF-α appear to correlate with the severity of the disease,² and circulating levels of the soluble TNF-α receptor 1 have been reported as a useful predictor of prognosis.³ Furthermore, elevated circulating inflammatory cytokines (including TNF-α, Figure 1) have been associated with myocardial damage and reduced pump function, suggesting they may contribute to the deterioration of the heart.⁴ It is now believed that increased levels of plasma pro-inflammatory cytokines are a result of an underlying immune process involved in the progression of CHF. In order to inhibit this process and prevent the progression of the disease, immunomodulatory agents (with a significant emphasis on TNF-α inhibition) have been studied as potential treatments for CHF. The results to date have been largely disappointing, with the possible exception of pentoxifylline (PTX)—a phosphodiesterase inhibitor with novel immunomodulatory properties.

Figure 1.

The theoretical effects of tumour necrosis factor-α (TNF-α) on the failing human heart. Damaged myocardium secretes tumour necrosis factor-α, leading to negative inotropy, apoptosis, and endothelial dysfunction. Tumour necrosis factor-α is secreted into the circulation, where a systemic immune response ensues. Leucocyte activation occurs, resulting in more tumour necrosis factor-α secretion. The net effect results in collateral myocardial death, which is presented symptomatically as chronic heart failure. PBMC, peripheral blood mononuclear cell.

The controversial role of tumour necrosis factor-α in chronic heart failure

Tumour necrosis factor-α is a pleiotropic pro-inflammatory cytokine which is secreted by a wide range of immune cells. Its function varies significantly depending on the immunological circumstance. Tumour necrosis factor-α is well described as having an important role in the immune response during infection and inflammation. At the level of the myocardium, cardiomyocytes have been documented to express both TNF-α and their corresponding receptors.⁵ Myocardial damage from endotoxin,⁶ ischaemia,⁷ and calcium overload⁸ is known to up-regulate synthesis of TNF-α, which results in a systemic increase in TNF-α concentration. However, whether TNF-α has direct deleterious effects on myocyte function/metabolism remains a contentious issue. A previous case report suggested that intravenous TNF-α administration was associated with the development of dilated cardiomyopathy.⁹ TNF-α has also been described to promote myocyte dysfunction by triggering apoptosis,¹⁰ uncoupling of B receptors from adenylate cyclase,¹¹ and promoting remodelling of the ventricle.⁴ This putative ability of TNF-α to damage the myocardium has frequently been reported as a mechanism through which the immune system could advance CHF. Nevertheless, alternative evidence exists to refute this evidence and even support a protective effect of TNF-α in CHF. Kurrelmeyer et al.¹² demonstrated in a murine model of ischaemic infarction that mice deficient in TNF receptors endured more extensive levels of apoptosis, in comparison with standard controls. They concluded that TNF signalling resulted in cytoprotective signals, which then prevented cardiac myocyte apoptosis after ischaemic infarction.

Possibly of more interest, are the results of clinical studies which have specifically looked at the effect of direct TNF-α inhibition. The RENEWAL trial¹³ was the combination of two large studies that assessed etanercept (a recombinant TNF receptor which functionally inactivates circulating TNF) in collectively over 2000 CHF patients, with ejection fractions <30% and NYHA class II–IV symptoms. Both contributing trials were stopped prematurely owing to a lack of observed benefit. Furthermore, in one of the trials, there was a suggestion of worsening clinical status with active treatment.

Infliximab, a monoclonal antibody to TNF-α, was also investigated in the ATTACH trial.¹⁴ In this study of 150 CHF patients with severe symptoms (NYHA III or IV) and ejection fractions <35%, participants were randomized to placebo or infliximab (two groups of either 5 or 10 mg/kg). Surprisingly, after a follow-up period of 28 weeks, the primary endpoint (death from any cause or hospitalization for heart failure) was significantly increased in the 10 mg group. There was no significant effect over placebo in the 5 mg group.

Pentoxifylline: alternative actions to tumour necrosis factor-α inhibition?

Pentoxifylline is a phosphodiesterase inhibitor which is commonly used in the treatment of peripheral vascular disease. A wide range of documented immunomodulatory properties have been described, including the down-regulation of TNF-α synthesis. Consequently, numerous in vitro and in vivo studies have explored the effects of TNF-α inhibition by PTX in an attempt to limit the reputed injurious effects of immune activation in CHF. In vivo experiments have specifically shown that PTX suppresses gene transcription of TNF-α,¹⁵ preventing its synthesis. In vivo PTX attenuates the increase of TNF-α in response to endotoxin⁶ and also attenuates levels of interleukin (IL)-12.¹⁶

Of significant importance, PTX is also documented to possess TNF-α-independent immunomodulatory effects. These may play an important role in the treatment of peripheral vascular disease and in theory may be translated into the treatment of CHF (Figure 2). PTX increases the flow and migration of peripheral blood mononuclear cells (PBMCs) through capillaries¹⁷ by reducing cell stiffness and increasing intracellular concentrations of cyclic nucleotides.¹⁸ PTX also inhibits the proliferation of PBMCs, attenuates IL-12 release,¹⁹ and prevents adherence to the cell matrix and the endothelium.^20,21 It has also been demonstrated that PTX can enhance the oxidative burst of PBMCs (which although contradictory to the proposed mode of action of PTX, would augment immune responses in vivo).²² However, of particular interest, PTX has been shown to inhibit T cell,²³ and NK cell²⁴ cytotoxicity, in addition to down-regulating IFN γ, GM-CSF (and TNF-α secretion).²³ In lymphocytes, its administration also attenuates the cell surface expression of the IL-2 receptor.²⁵ Furthermore, it has been shown that complete inhibition of intracellular adhesion molecule-1 expression and the production of IL-8 and CCL2 by cytokine-activated epithelial cells occur in the presence of PTX.^26,27

Figure 2.

Immunomodulatory mechanisms of pentoxifylline. Pentoxifylline has many reported immunomodulatory effects. These include immune cell effects such as the prevention of leucocyte proliferation and the inhibition of direct (perforin/granzyme) and indirect (cytokine production) cytotoxic responses. Pentoxifylline also modulates the endothelium via the down-regulation of cell surface adhesion molecule expression and chemokine production, and also via moderation of apoptotic pathways. PBMC, peripheral blood mononuclear cell.

During CHF, excessive cell apoptosis leads to structural abnormalities in the myocardium.^28,29 Cardiac cell apoptosis occurs only under stress, which may be due to ischaemic injury or pressure overload. Using a gene knockout mouse model, it has been reported that apoptosis occurs during the onset of CHF following biochemical stress.²⁸ Circulating levels of the cell-surface apoptosis signalling/death receptor Fas ligand(L) are also raised in cardiomyopathy.³⁰ Therefore, apoptosis is likely to play a pivotal role in disease progression. Pentoxifylline inhibits apoptosis pathways in cultured PBMCs and significantly reduces circulating levels of FasL, suggesting a possible mechanism of action may be in the inhibition of apoptosis.^31,32 By preventing cell death, PTX might therefore preserve the myocardium, protecting it from immune-mediated apoptosis and subsequently delaying the progression of CHF.

The endogenous stress signals provided by damaged (apoptotic) myocardium may be responsible for the systemic and localized immune activation observed in CHF. This is characterized via elevated levels of circulating and myocardial inflammatory cytokines in these patients. Therefore, a beneficial effect of PTX in CHF may be the prevention of myocardial damage via the suppression of apoptotic processes, rather than the suppression of TNF-α, which although elevated, does not appear to have a clearly defined role in the pathophysiology of CHF. Indeed, TNF-α may represent a bystander molecule, and its characteristic elevation could merely be representative of general immune activation. The clinical PTX studies have provided further evidence to support this.

Pentoxifylline in clinical trials

In a cohort of patients with coronary artery disease (n = 64), PTX therapy (6 month placebo-controlled, double-blind study) was reported to reduce circulating levels of C-reactive protein and TNF-α.³³ The study also reported that PTX patients had an increase in the anti-inflammatory cytokine, TGF-β1. The authors concluded that PTX has a beneficial effect on the immune system via the immunomodulation of ‘inflammatory activity’, and may be of therapeutic use. Pentoxifylline was also investigated in a series of small-scale clinical trials in patients with CHF. The promising results of these trials have generally supported the theory that there is an immune-mediated pathophysiology in CHF. However, the exact mechanism behind the benefits of PTX in heart failure remains a subject of significant interest. Conflicting reports exist on whether it modulates the immune system by inhibiting TNF-α synthesis or through alternative processes.

The first double-blind, placebo-controlled trial of PTX in CHF was conducted by Sliwa et al.³⁴ This group selected 28 patients with stable NYHA functional class II or III CHF of unknown aetiology. The two groups of patients (both n = 14) were randomly assigned to PTX 400 mg three times daily or a matching placebo. The study reported that after 6 months, there was a significant improvement in both the distribution of NYHA functional class and left ventricular ejection fraction (LVEF) in the treatment group (mean 38.7 vs. 26.8%, P < 0.01 and P = 0.04, respectively). The investigators also found that circulating TNF-α concentration was lower in the treatment group in comparison with the placebo group (P = 0.001). This study was the first to link TNF-α reduction to clinical improvement in CHF patients.

A similar study in 38 ischaemic cardiomyopathy patients demonstrated that PTX treatment resulted in a significant improvement in functional class and LVEF (22.6 vs. 36.8%, P < 0.005 and P < 0.05, respectively).³¹ This study also found that PTX therapy resulted in a significant reduction in soluble Fas receptor, TNF-α, and C-reactive protein concentration. There was a significant reduction in TNF-α from baseline to 6 months in the treatment arm compared with the placebo. These encouraging results suggest that improvements in LVEF and NYHA classification are associated with a reduced plasma levels of TNF-α and Fas receptors.

Skudicky et al.³⁵ (as part of the same group of investigators) conducted an additional study of 49 patients with the same inclusion criteria as Sliwa et al. They also reported a reduction in Fas receptor concentration, which correlated with the degree of improvement in left ventricular function in patients with cardiomyopathy. There was no significant correlation between alteration in TNF-α concentration and changes in ejection fraction. Again, there was a significant reduction in plasma TNF-α; however, this occurred in both patients whose clinical status improved and in those patients who did not improve. This study (which was the largest cohort the group reported on) therefore suggests that the beneficial effects of PTX treatment may not be instigated via TNF-α down-regulation, but by other TNF-α-independent mechanisms. In addition, the same investigators conducted a similar study in 39 patients treated with carvedilol and ACE-inhibitors.³⁶ Patients were randomized to PTX 400 mg three times daily (n = 20) or a placebo (n = 19). In the treatment group, 66.6% of patients had an improved NYHA functional class compared with just 10% of patients in the placebo group (P = 0.01). There was also a significant improvement in the mean change in LVEF from baseline to 6 months in comparison with the placebo group (25 vs. 31%, P = 0.04). It was also reported that there was no significant alteration in the concentration of TNF-α in patients on PTX. Instead, the investigators showed that PTX treatment caused a significant decline in Fas receptor levels compared with the placebo group.

A further report from Sliwa et al.³⁷ detailed the effects of PTX use in a cohort of peripartum cardiomyopathy patients (n = 59). The study found that both NYHA functional class and LVEF significantly improved in the treatment group (∼23 vs. 44%, P = 0.006 and P = 0.01, respectively) after only 6 months of follow-up. No significant reduction in TNF-α in comparison with the placebo group was observed.

Collectively, these trials demonstrate that PTX therapy in CHF appears to result in an improvement in clinical symptoms, yet not necessarily in co-association with a reduction of circulating TNF-α concentration. However, it should be noted that a study in 2004 by Bahrmann et al.³⁸ also reported the effects of PTX in a group of CHF patients (n = 47). Interestingly, this study failed to report a beneficial effect of active treatment on any clinical variable, or TNF-α and IL-6 concentration. This suggests that further clinical studies are required on a larger scale to fully determine the extent of clinical benefit from PTX in CHF.

Additional vascular effects of pentoxifylline

Pentoxifylline is traditionally used as a treatment for peripheral arterial disease. The exact mechanism behind its clinical benefits in this setting is also poorly defined, but has been attributed to its ability to affect cellular blood flow through the microvasculature. This appears to primarily occur through reductions in whole blood viscosity in addition to improving red cell deformation.^39,40 However, there is also evidence that PTX may induce arterial vasodilatation, although at doses higher than those of standard oral dose treatment.⁴¹ Clearly, this latter property may also have relevance to CHF, given other vasodilatory agents have been documented to offer beneficial effects on mortality and morbidity in this setting.⁴²

Discussion

There is emerging evidence that the immune system is involved in the progression of CHF. Both in vitro and in vivo studies have demonstrated that PTX has immunomodulatory effects and may play a role in preventing disease progression. The clinical PTX trials have potentially exposed a fallacy in our understanding of how it benefits patients with CHF. The initial rationale for using PTX as a therapy for CHF was due to its ability in preventing TNF-α transcription, which theoretically would prevent cytokine-mediated damage to the myocardium. Yet, the majority of clinical trials to date have reported improvements in clinical outcome following PTX treatment, without a concordant effect on TNF-α modulation. Whether this is representative of a lack of involvement of TNF-α in the pathogenesis of heart failure is unclear. However, TNF-α is a multifaceted cytokine that is raised in response to multiple physiological processes and forms part of the intrinsic stress response to myocardial injury. Moreover, despite several previous attempts at direct TNF-α modulation, no benefit has been seen from this approach. Therefore, we do not fully understand the complex role (if any) TNF-α may play in heart failure. What we can conclude is that PTX has shown significant early promise as a novel new treatment in CHF, though in clinical trials with small study groups. The conferred benefits may occur through subtle effects on arterial vasculature, through complex TNF-α-independent immunomodulatory effects, or a combination of both. In order to determine whether it fulfils this early promise, more clinical trials on a larger scale are required.

Conflict of interest: none declared.