A role for cytokine measurement in therapeutic monitoring of immunosuppressive drugs following lung transplantation

The strength of the allo-immune response to a transplanted organ [1,2] currently mandates combined pharmacological immunosuppression to prevent rejection. The calcineurine inhibitor Cyclosporine remains an important component of immunosuppression regimens post lung transplantation with the International Registry of Heart and Lung Transplantation recording 50% of patients receiving this drug.

Cyclosporine requires therapeutic drug monitoring with the aim of maximizing efficacy whilst minimizing the risk of toxicity. Historically the trough (predose or Co) level has been used to guide the dose of cyclosporin and the dose range in lung transplantation has been determined by rather empirical review of studies in kidney transplantation [3]. A survey conducted in 1994–95 showed that most lung transplant centres however, recommended a higher range of target concentrations during the early post operative period [4].

The recommended therapeutic blood concentrations however, are broad and may vary according to the clinical situation and the choice of other immunosuppressive agents being administered. Relatively recent data has supported the use of therapeutic drug monitoring of cyclosporin by the use of whole blood drawn two hours after administration of the dose (C2) with data showing a better prediction of clinical effect [5]. Nevertheless therapeutic drug monitoring has certain general limitations with the concentration measured in blood, not necessarily the concentration of the drug at the site of action. The alternative approach, as discussed in a study by Hodge and colleagues in the January 2005 issue of Clinical and Experimental Immunology is to measure the therapeutic action of the drug either directly or indirectly [6].

In general the approaches involve in vitro assessment of T cell activation by measuring cytokine release from peripheral circulating cells either in the resting or stimulated state [7–9]. Whilst these techniques have great potential, it is not yet established whether they are predictive of clinical events. Moreover the complexity of the allo-immune response may make it difficult to model using such assays. There is also a problem regarding which compartment of the immune system should one study in order to give the best biological information. Most studies use peripheral circulating cells rather than intragraft immune cells which may be more relevant in terms of effecting graft injury.

In a broad sense, the process of immune cell-mediated allograft rejection is regulated by an array of cytokines. Some of these, for example the chemokines, are produced by stressed graft cells and activated immune cells whilst others, such as IL-2 and IL-4, are produced by activated immune cells. Given the importance of this range of cytokines, it is perhaps surprising that immunosuppressive calcineurin inhibitors such as Cyclosporin A (Cs-A) are effective as they directly regulate the expression of only a very limited range of cytokines. This includes blockade of de novo expression of the T cell growth factor IL-2 following the activation of resting T cells [10] and up-regulation of the expression of TGF-β by a range of parenchymal cells [11]. Clearly the analysis of cytokine expression has the potential to provide clinically useful information about the state of allograft rejection or acceptance. However, a number of issues must first be addressed, including:

Which are the most useful cytokines to examine?

In many cases the expression of an individual cytokine can be quite transient and occurs in association with the development of a specific pathology. For example, CXCL8 (IL-8) is produced at an early stage during transplantation [12]. Indeed, this chemokine can be induced before transplantation by the process of brain death associated with organ donation. The major targets of this cytokine are polymorphonuclear leucocytes which infiltrate the graft during the innate inflammatory reaction associated with transplantation; this ‘reimplantation response’ may contribute to the primary graft nonfunction which is a major cause of recipient morbidity [13].

Following transplantation it is clear that the expression of cytokines such as IL-2 and IFN-γ are associated with the acute inflammatory events associated with acute rejection. However, the expression of IL-2 is necessarily transient, which limits its effectiveness as a marker of rejection; indeed, prolonged expression of this cytokine can drive allospecific T cells to premature apoptosis through activation-induced cell death [14]. The prognostic value of the detection of IL-2 is also limited by the potential of activated T cells to respond to graft-cell produced growth factor such as IL-15, which is produced constitutively even in the presence of Cs-A [15]. The detection of IFN-γ provides a more robust measure of specific activation of the T cells involved in potentially graft-damaging Th1 immune processes.

The chronic stages of allograft fibrogenesis and failure are dominated by the expression of growth factors such as TGF-β[16]. However, simple detection or measurement of TGF-β is of very limited value since it is only the activated form of the cytokine that is likely to have any graft damaging activity. It is also important to bear in mind the immunomodulatory activity of this cytokine, which has been implicated in the activity of at least some regulatory T cells [17]. Hence, a single factor might simultaneously have beneficial and damaging activities at different stages in the development of allograft pathology.

Where should cytokine expression be assessed?

There are two potential sites from which cells or fluid can be drawn for cytokine analysis; these are peripheral to the graft, including the blood and the spleen, and within the allograft itself. It seems clear that acute rejection is promoted by initial antigen presentation by donor cells which have migrated to the recipient patient's spleen. Local T cell expansion can then occur, allowing cells to traffic around the body until they are recruited by activated graft endothelium. In addition to this simple model, it is known that some allospecific cells are also activated directly within the graft tissues.

For routine diagnostic purposes, assays based on peripheral blood would be advantageous. However, most cytokines bind to proteoglycan components of the cell surface or extracellular matrix [18] and have a very limited availability within the blood, effectively limiting the prognostic value of any assay of soluble cytokine proteins. However, a range of approaches have been developed which allow evaluation of cytokine production by peripheral blood-borne lymphocytes. The most powerful of these approaches allow direct scrutiny of the activity of the donor-antigen specific T cell subpopulation.

It is also possible to examine intragraft immune and parenchymal cell expression of cytokines using diagnostic tissue retrieved from a transplanted organ. As direct examination of biopsy tissue still provides the ‘gold-standard’ measure of allograft rejection, it is likely that the analysis of cytokine expression within the graft will be of the greatest diagnostic and prognostic value.

How can cytokine expression be measured?

Cytokine levels in whole blood can be assayed easily by ELISA or by application of more recent multiparameter bead based assays; however, as stated above, the value of such assays remains debateable. The production of cytokines by individual immune cell types can be assessed using a range of techniques. Of crucial importance for these assays is the requirement for cell activation prior to analysis. In the case of T cells this can be accomplished by polyclonal stimulation with mitogens active on the cell surface or within the cell or by antigen-specific stimulation. In the case of alloreactive cells, when the frequency of the responding cells is relatively high, activation by alloantigen presentation can provide a useful technique to assess the potential to mount a productive antigraft immune response. Following cell activation suitable analytical readout systems include ELISPOT [19] or limiting dilution analysis [20], which both provide sensitive frequency values, or flow cytometric techniques to investigate the phenotype of those cells which are stimulated to produce intracellular cytokines such as IFN-γ or the Th2 cytokine, IL-4. In order to assess the efficacy of immunosuppression, these assays can be performed in the presence of agents such as Cs-A.

Intragraft cytokine expression can be measured in three ways. Firstly by assay of mRNA, secondly by measurement of protein expression and thirdly by assay of the expression of robust markers which themselves are induced by response to a specific cytokine. In each case detection could be achieved using disrupted tissue, but this seriously limits the acquisition of spatial information. The most sensitive techniques now include laser microdissection for tissue-specific real-time quantitative PCR, in situ hybridization and immunocytochemistry. Whilst in situ hybridization is often limited by the transient nature and low level of cytokine mRNA expression [21], immunocytochemistry is aided by the potential of cytokines to bind to proteoglycans within allograft tissues; importantly, modern image analysis of immunofluorescence signals can allow partial quantification of localized immunochemical labelling [22]. Examples of indirect cytokine detection within allograft tissues include the induction of class II MHC antigen expression following exposure of parenchymal cells to local IFN-γ and the expression of CD103 by TGF-α stimulated T cells [23].

It is clear that no one approach to cytokine analysis is sufficient to produce simple diagnostic or prognostic data which has validity at every stage during the rejection of a transplanted organ. However, combination of results from appropriate cytokine analysis with conventional histological data derived by investigation of biopsy specimens will certainly increase our understanding of the mechanisms which cause tissue damage following organ transplantation.

The potential role of monitoring immunosuppression using biological markers as opposed to pharmacokinetic ones is therefore very complex. Though exciting, it requires a good deal of further research before it is likely to become of practical value in clinical lung transplantation.