“This modeling thing, it’s pretty easy, but actually it’s also really tough.” Cara Delevingne

The necessity

With the ultimate goal to reduce mortality/morbidity in patients, animal modeling of diseases has been limited by poor translation [1, 2]. This is often fueled by the low fidelity of available model systems [3, 4], their inappropriate study designs [2] and selective use of animal data [5, 6]. When compared to other inflammatory states (e.g. arthritis, atherosclerosis), the complexity of sepsis has hampered the development of high-fidelity models. However, this challenge can be aptly embraced by building on recent advances in the understanding of sepsis pathophysiology and avoiding past errors. Any promising sepsis model must be (a) specifically tailored to the posited hypothesis, (b) “reverse translated” to its clinical counterpart [7, 8] and (c) adjusted as new pathophysiological evidence emerges. This is echoed by the US Food and Drug Administration (FDA) in their 2010 Guidance for Industry and FDA Staff: “FDA believes that the animal…(model)…should provide a test system that offers a best attempt at simulating the clinical setting”. (General Considerations for Animal Studies for Cardiovascular Devices; http://www.fda.gov).

Unfortunately, while the clinical definition of sepsis is currently in its third iteration [9] and the Surviving Sepsis Campaign (SSC) Guidelines for patient management have been updated three times [10], pre-clinical sepsis research has not been subjected to any organized attempt at introducing best practices, management guidelines and standardization [11]. This creates a large quality gap and confusion with conflicting data reflecting huge variations in, for example, insult severity, fluid resuscitation and study duration. Effective animal modeling and reporting guidelines have recently been proposed for other specific diseases such as pulmonary fibrosis [12], stroke [13, 14], heart failure [15] and malaria [16] making the void in the field of pre-clinical sepsis even more apparent. It is essential that animal models of sepsis continue to evolve. Lack of sufficient standardization of preclinical models will continue to limit the utility of sepsis animal research as a useful platform for advancing clinical outcomes and care in sepsis [17, 18] and will reduce the opportunities to identify and test new therapies.

The action

To address this perceived deficit, the Ludwig Boltzmann Institute of Experimental and Clinical Traumatology in the AUVA Research Center organized in May 2017 in Vienna a Wiggers-Bernard Conference on “Pre-clinical Modeling in Sepsis: Exchanging Opinions and Forming Recommendations”. The key goal was to create publishable material that characterizes elements that should be included in pre-clinical sepsis studies and defined by the so called “Minimum Quality Threshold in Pre-Clinical Sepsis Studies” (MQTiPSS) descriptor. The Wiggers-Bernard Conference participants identified and addressed several broad, critically-important concepts in animal sepsis modeling. A total of 31 experts from 13 countries participated in the initiative (including five members of the Sepsis-3 definitions task force) and were divided into 6 thematic Working Groups: (1) study design, (2) humane endpoints, (3) infection types, (4) organ failure/dysfunction, (5) critical fluid resuscitation and (6) antimicrobial therapy.

The initiative consisted of three phases: (a) preparatory (prior to the meeting; approximately 3 months), during which participants performed a systematic review of the 260 top cited (over 29,000 citations in aggregate) 2003–2012 pre-clinical publications (using ISI Web of Knowledge database; query: “sepsis model”; 374 individual experiments analyzed) and identified the key modeling topics to be discussed, (b) discussion during which the participants spent 2 days at the Wiggers-Bernard Conference examining pre-clinical sepsis models and ultimately voted to reach consensus on the proposed points (either at the “recommendation” or “consideration” strength), and (c) post-meeting refinement of the accepted points and finalization of the arguments to be included in the final publications (using a modified Delphi method; approximately 3 months). Following the format used by the Sepsis-3 task force [8], at least 2/3 (over 65%) of the votes were required for approval of a proposed point.

The proposed outcome

First, a definition for an animal model of sepsis was formulated and (unanimously) approved: “An experimental animal (mammal) model of sepsis should be defined as life-threatening organ dysfunction caused by a dysregulated host response to an infection.” Second, Wiggers-Bernard Conference participants reached consensus on 29 points; 20 at “recommendation” strength and 9 at “consideration” strength (listed in Tables 1, 2 and 3). All consensus points were reached either unanimously or with no more than 2 abstentions per point (point 8). The “recommendation” strength indicates virtually unanimous agreement among the 31 participants, regarding both the content and the need for rapid implementation. Issues that require additional discussion before final recommendations could be made were classified as considerations.

The current executive summary briefly describes the Wiggers–Bernard Conference initiative and presents the compiled consensus points. The details of the recommendations/considerations are published in three separate papers [19,20,21] subsequently appearing in the 2019 January issue of the Shock journal. The Tables 1, 2 and 3 summarize the main MQTiPSS consensus points published in those articles: Part I—Table 1 [19], Part II – Table 2 [20] and Part III—Table 3 [21]. Each publication is built on two (related) Working Group themes and includes a narrative clarifying caveats and intricacies related to the accepted consensus points.

Table 1 Combined Recommendations and Considerations from the Working Group (WG) 1 and 2
Full size table
Table 2 Combined Recommendations and Considerations from the Working Group (WG) 3 and 4
Full size table
Table 3 Combined Recommendations and Considerations from the Working Group (WG) 5 and 6
Full size table

The future

The presented consensus has not received formal endorsement from professional bodies. Writing an initial consensus was a strategic decision given that an expert opinion report has a shorter publication turnaround and our intention was to rapidly introduce the MQTiPSS concept. The Wiggers–Bernard Conference was conceived not as a one-time event but rather as an initial “call-to-arms”; an invitation to interested parties to provide further refinement and expansion of the proposed points. The on-going expansion initiatives include formation of a Task Force (under the auspices of the Shock Society; June 2017) for creation of robust, defined parameters to score sepsis models for clinical relevance. Another iteration of the Wiggers-Bernard Conference on animal sepsis models is planned for October 2019 at the joint conference of the European Shock Society and International Federation of Shock Societies in Crete, Greece.

In summary, we believe that the proposed guidelines represent the first concrete steps toward creation of a realistic framework for standardization of animal models of sepsis (i.e. MQTiPSS). Such a framework, once widely employed, will improve the quality of pre-clinical investigation and arm clinicians with better tools for combating sepsis in patients.