Models of Human AKI: Resemblance, Reproducibility, and Return on Investment

Extensively studied for several decades, animal models of human AKI have, to date, yielded no specific therapy that benefits the human disease, either in preventing its occurrence, ameliorating its severity, hastening its recovery, or retarding the risk of ensuing CKD. Such inability to translate salutary strategies uncovered in these models into therapies for human AKI has led to a questioning of the value of these models: specifically, their resemblance and fidelity to human AKI; the reproducibility of findings so derived; and the likelihood that financial and other resources supporting this investigative approach would eventually receive a return that benefits human AKI. Overarching this kidney-centered discussion are broader ones in relevant biomedical communities, national as well as international, regarding a concerning relative lack of congruence of conclusions in preclinical studies and the slow pace of translating findings obtained in animal models into new therapies for human disease.^1–5

In the current issue of the Journal of the American Society of Nephrology, de Caestecker et al.⁶ advance this discussion by advocating attention to specific potentially addressable issues in the design and conduct of preclinical studies in AKI. To frame their analysis, the authors chose two complementary examples to illustrate salient considerations. The first example pertains to the long and continuing clinical search for evidence regarding the efficacy of sodium bicarbonate and/or N-acetylcysteine in clinical contrast-induced AKI; clinical evidence thus far remains no more than a will-o’-the-wisp. As de Caestecker et al.⁶ point out, this clinical search was weakly founded and may have been prematurely undertaken since the supporting preclinical evidence was derived from a limited number of studies in which the investigative approach was less than rigorous and the conclusions insufficiently persuasive. In contrast, and as discussed in the second example by de Caestecker et al.,⁶ the preclinical evidence demonstrating the efficacy of erythropoietin in rodent models of AKI was strong and supported by the majority of studies that addressed this issue; disappointingly, translational studies failed to conclusively demonstrate a beneficial effect of erythropoietin in human AKI.⁶

Drawing upon these examples and an in-depth analysis of the relevant literature, de Caestecker et al.⁶ propose a major re-engineering in the design and reporting of preclinical studies and a widening of the compass for preclinical study protocols. Their proposals provide, among others, the following specific recommendations: robust statistical methods that include adequately powered sample size; experimental models utilized in ways that incorporate the salient clinical conditions surrounding human AKI; attentiveness to the effects of gender, age, and genetic heterogeneity on AKI; randomization and a blinded approach; comprehensive pharmacologic studies; and the publication of negative studies.⁶ This timely and important contribution by de Caestecker et al., thus, brings to the AKI field recommendations proposed and considered elsewhere, and ones that will promote in AKI research both rigor and reproducibility, the latter representing the bedrock of any scientific field.

These recommendations of de Caestecker et al.⁶ are provided under the rubric of guidelines, albeit ones that are preliminary. Guidelines, once endorsed, have a propensity to slide almost imperceptibly from recommendation to expectation and onto requirement; their consequences should thus be carefully considered. As noted by de Caestecker et al.,⁶ these guidelines will impose substantial additional costs, the need to recruit relevant resources and expertise, and more administrative layering in the conduct of preclinical research. Securing the additional financial and other support needed to meet these guidelines may be challenging, especially in view of the inclemency of the current research climate: specifically, the community that primarily conducts preclinical research (physician-scientists and basic scientists) and the institutions that mainly support such research (academic health centers and university science departments) all exist not only in troubled times, but also in an unnerving predicament of an uncertain future^7,8; in this milieu, institutional support for unfunded research continues its decline. Without readily obtainable financial and other resources needed to meet these guidelines, a skeptic may even raise the specter of declining appeal for scientists to pursue preclinical investigation, with an attendant attrition in the size and functionality of the research community dedicated to such studies. Such a skeptic may thus pose the following question: Would the current level of functionality of the preclinical research community, however imperfect, be preferable to one that becomes contracted and thus less capable in undertaking preclinical studies?

de Caestecker et al.⁶ anticipate the challenges posed by these guidelines and suggest ways in which added costs can be curtailed and logistical and resource-related issues can be addressed; additionally, statistical outcomes and approaches may be applied that mitigate the increase in sample size that may be otherwise required. Importantly, de Caestecker et al.⁶ recommend that stakeholders with a vested interest in preclinical studies—scientists who conduct such studies, the pharmaceutical industry which peruses such studies for leads into drug development, clinical trialists, funding agencies, and the Food and Drug Administration, among others—should engage in an ongoing dialogue regarding reproducibility and strategies intended to improve it, including those strategies outlined above. Such a dialogue should broach germane issues, including the tendency of scientists to leave their negative findings unpublished and the inclination of pharmaceutical companies to relinquish support too hastily for potential therapies that merit further study.

A major impetus for these guidelines is the recognition that “…our current strategies have failed to deliver new therapies…”⁶ Relevant considerations include the general manner in which biomedical discoveries tend to occur and the duration of investigation before failure to deliver new therapies may be declared. The pathophysiology of AKI has been studied and therapeutic strategies have been pursued in preclinical models for some 50 years; in other areas of nephrology that are the beneficiaries of novel therapies, the journey leading to these therapies is even longer. For example, the introduction of angiotensin–converting enzyme inhibitors and angiotensin receptor blockers to retard the progression of CKD, and the utilization of recombinant erythropoietin to treat the anemia of ESRD/CKD, had their respective origins in preclinical observations undertaken approximately a century previously: the discovery of renin by Tigerstedt and Bergman in 1898 in the case of angiotensin–converting enzyme inhibitors/angiotensin receptor blockers and the demonstration by Viault in 1890 that high altitude stimulates erythrocytosis in the case of erythropoietin.^9,10 Thus, at least as precedented by these two examples of novel therapies in nephrology, investigative journeys culminating in therapeutic success are usually long and fitful, and the felicitous outcomes are rarely reached in a predictable and prosaic way. The path of scientific work and progress, to quote Robin Marantz Henig, tends to be “mysterious, meandering, and messy.”¹¹ However, an improbable occurrence or a chance finding may lead to a quickening of the step and the arrival at somewhere unexpected, with new directions now disclosing themselves; serendipity and “black swans” point the way to biomedical breakthroughs¹² as do, in certain instances, “brilliant blunders.”¹³ Greater rigor in preclinical study design in AKI will lead to more reproducible and reliable information. Reproducibility, however, is necessary but not sufficient for biomedical breakthroughs, and any new therapy for AKI would be, indubitably, one such breakthrough. Thus, just as important would be if, with this surer and more secure knowledge base, there is a greater perceptivity in recognizing and capitalizing on those black swans and serendipitous findings that may prove critical for therapeutic innovation in human AKI.

Preclinical studies have contributed substantially to therapies introduced to and now established in clinical practice. In searching for new therapies in a given field, the contribution of employed preclinical models can always be increased by improving the resemblance of these models to human disease, the extent to which these models recapitulate the clinical setting, and the reliability of information so derived. By delineating potentially addressable issues in the design, implementation, and reporting of preclinical studies in AKI, the guidelines by de Caestecker et al.⁶ will promote the reproducibility of these studies and, hopefully, their return on investment, the latter culminating, ideally, in novel therapies for human AKI. It must be underscored, however, that initiatives with new expectations require enabling resources. It is thus essential that requisite financial and other resources be made readily available to investigators if the raison d'être of such guidelines is to be met, their intent delivered, and their hope fulfilled.

Disclosures

None.