Interleukin-10 gene down-expression in circulating mononuclear cells during infusion of drotrecogin-α activated: a pilot study

To improve the outcome, continuous infusion of drotrecogin-α activated (DAA) is recommended for over 96 h at a rate of 24 μg/kg/h in septic shock patients as early as possible [1]. This dosage has raised cost/effectiveness concerns. Yet, since DAA has been made commercially available, in vitro studies have highlighted many more properties for this molecule and it can no longer be ignored [2]. A convenient biomarker of its best use would be welcome to select patients who could truly benefit from this treatment. This test has never been done for several reasons: (1) single and serial measurements of plasma concentrations of inflammation biomarkers are inconsistent and do not reliably predict outcome in septic patients [3], and (2) since the effect of infused DAA probably changes according to the endogenous concentration of activated protein C at any given moment, it is currently difficult to know if patients are responding to this drug and whether DAA dosage should be adapted when treatment is started without monitoring its plasma concentration over 96 h, which is a complicated and expensive procedure.

In addition, there are methodological pitfalls in septic shock trials relating to the diversity of sources of infection and the exact time the infection started. Also, the inflammatory response triggered by pathogen-associated molecular patterns through Toll-like receptors (TLR) activates a fast response by the nuclear factor kB (NF-kB) pathway. Because the cellular signaling of TLRs can be modified by polymorphism [4] and the subunits of NF-kB expression may be affected by DAA [5], participants in DAA clinical trials should have TLRs and NF-kB expressions as similar as possible so that groups can be appropriately compared. Only then will genes targeted by the NF-kB complex be expressed on a sound basis of comparison. Although, conclusions would be more reliable, large clinical trials taking all these parameters into account would be much too expensive.

This study was conducted in real-life conditions of septic shock management to address the following question: does DAA have any tangible effect on the early pro-inflammatory response to septic shock, measured as a change in TNF-alpha (TNF-α), interferon-gamma (INF-γ), and interleukin-10 (IL-10) mRNA expressions in circulating mononuclear cells (CMNC) harvested from patients and true controls? If so, this would provide an early indication of improvement, and perhaps a DAA efficiency biomarker.

This study was approved by our institutional review board for human research and informed written consent was obtained from each participant. Patients with inherited or acquired immunodeficiency were excluded.

Over one year, we included 16 consecutive patients with septic shock [6] treated with DAA (DAA+ group) at the standard dosage (24 μg/kg/h for 96 h), resulting in a mean plasma DAA concentration of 45 ng/mL [2]), and 16 patients as a control group fulfilling septic shock criteria and eligible for DAA but contraindicated (platelet count < 30,000/mm³ and/or uncontrolled bleeding).

We defined T1 as the moment within 12 h after fulfilling septic shock criteria and before giving DAA; T2 was set 36 h later. The Simplified Acute Physiology Score II [7], the Logistic Organ Dysfunction score [8] and 28-day mortality were recorded.

Plasma concentrations of IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, TNF-α, IL-10, IFN-γ and IL-12 were measured (CBA Human inflammation kit, BD Biosciences, San José, CA, USA) and lympho-monocytes subpopulations determined by flow cytometry (Becton Dickinson, Rungis, France).

Aliquots of 20 ml of whole blood were drawn in EDTA and, CMNC were isolated by Ficoll gradient centrifugation (Eurobio, Les Ulis, France) to obtain aliquots of 10⁶ cells. RNA extraction and reverse transcription and quantitative real-time PCR were performed with a gene reporter (hydroxymethyl-bilane synthase, HMBS) as reported [9]. The primers for NF-kB sub-units (p50, p65 and IkBa) were purchased from Genome Express, Montreuil, France. For cytokines transcriptional expression (TNF-α, IL-10 and IFN-γ) QuantiTect Primer Assays primers (Qiagen, Courtaboeuf, France) were used. Gene expression was assessed using the 2-DDCT method as reported [10].

The TLR-2 (G2408A) and TLR-4 (A12874G and C13174T) polymorphisms were detected using a hybridization probe assay according to the reported method [4].

Increased plasma concentration of cytokines such as IFN-γ, TNF-α or IL-10 is a hallmark of septic shock and has negative consequences for recovery, although no clinically relevant thresholds have been proposed [3, 12]. However, data on plasma cytokine concentrations do not seem to determine outcome because the exact onset of sepsis-driven cytokines release is usually not clearly definable in septic patients [13]. Therefore, in this study we checked cytokine gene expression by CMNC, and confirm that changes to them occur without consequences on plasma levels in accordance with previous studies [14]. The most significant drawbacks of this approach are the routine availability of the RT-qPCR technique in a hospital laboratory and the time required to obtain the lab test result, which nevertheless takes a few hours (less than six hours) with commercially available techniques of molecular biology.

Our data are in partial accordance with in vitro studies that claim that DAA up-regulates IL-10 in LPS-stimulated human monocytes [15]. Yet in vitro the DAA concentration required is significantly higher (120 ng/mL) than that achieved in vivo with standard infusion (45 to 52 ng/mL) [2, 16]. In our setting, IL-10 gene expression was dramatically increased in the non-surviving patients given DAA suggesting either a greater IL-10 gene up-regulation by DAA in patients with poor prognosis, or simply that DAA has no effect on high IL-10 up-regulation by sepsis itself. Since the baseline gene expression levels were not significantly different between groups whatever the outcome and treatment, we suggest that DAA has an influence that is not necessarily powerful.

One of the limitations of our study is that it was underpowered but our hypothesis can not be tested in a random fashion in a large population of patients since DAA administration can not be declined in the absence of contraindication for ethical reasons. Furthermore, in septic shock DAA may actually require a higher infusion rate than the standard one to reach cellular efficiency (that is, dampening IL-10 gene expression). Conversely, DAA may have no effect on patients with continuous IL-10 expression. Hence, maintaining DAA for 96 h as recommended would be questionable. In both situations, the level of IL-10 up regulation in CMNC after 36 h DAA infusion may become a biomarker of efficiency. The time required to obtain the IL-10 gene expression ratio is technically feasible during the working hours of the laboratory: consequently, in a patient infused for 36 h with a T2 test indicating a lack of IL-10 and IFN-γ gene down-regulation, the standard DAA dosage could be reduced after approximately half the scheduled dosage has been infused (which would result in a 50% saving of the drug). This warrants confirmation by a prospective randomized placebo study including only patients with IL-10 over expression.

If IL-10 and/or IFN-γ gene down-expression fail to occur despite 36 h of DAA infusion in septic shock patients, maintaining treatment for 96 h as recommended becomes questionable. IL-10 and IFN-γ gene expressions by CMNC could become a biomarker of DAA efficiency in this setting.