Letter to the Editor: The Value of a Paired Design Approach as a Strategy for Reduction of Animal Use

Dear Editor,

The Syrian hamster (Mesocricetus auratus) has been widely employed in experimental leptospirosis due to its high susceptibility and its ability to reproduce key clinical and immunologic features of the disease.¹ This well-established use naturally raises interest in how the model can be further explored to address specific translational questions. In this context, the hamster emerges as a promising experimental system for investigating the reproductive effects of leptospiral infection relevant to bovine disease, offering practical advantages in terms of management, cost, and experimental reproducibility.

Bovine genital leptospirosis (BGL) represents a chronic and frequently subclinical manifestation of infection caused by host-adapted leptospires, characterized by persistent genital colonization and associated with significant reproductive losses, including abortion, embryonic mortality, stillbirths, and weak calves.² Given its economic impact and the intrinsic limitations of conducting controlled experimental infections in cattle, the use of animal models remains essential for elucidating pathophysiological mechanisms and for evaluating therapeutic and vaccine strategies under standardized conditions.

In previous studies, our group demonstrated comparable outcomes of experimental infection and uterine colonization in hamsters infected with leptospires of the Sejroe serogroup, the primary serogroup implicated in BGL.³ We further validated the hamster as a model for acute clinical leptospirosis⁴ and for the asymptomatic genital form of the disease, characterized by sustained uterine infection in the absence of overt clinical signs.⁵ Within this framework, evaluating whether leptospiral infection affects reproductive performance, specifically birth rate, represents a critical step in mimicking bovine reproductive alterations and strengthening the validation of the hamster as an experimental model for BGL. Accordingly, we designed a study comparing reproductive outcomes in hamsters infected and uninfected with Leptospira santarosai, serogroup Sejroe, serovar Guaricura.

As expected for studies of this nature, ethical considerations were integral to the experimental planning. While the 3Rs framework (Replacement, Reduction, Refinement) is a well-established reference in animal research, its practical implementation often depends on context. In infectious disease and reproductive studies, Replacement options remain limited, rendering in vivo models indispensable. In such scenarios, Refinement assumes a central role, not only in terms of clinical procedures, animal handling, anesthesia, and analgesia, but also through improvements in experimental design and analytical strategies. As emphasized by the ARRIVE guidelines, rigorous design, appropriate statistical planning, transparent sample-size calculation, and comprehensive reporting are themselves refinement tools, as they reduce variability, prevent poorly designed studies, and avoid needless repetition.⁶

Although Reduction is frequently cited as an ethical objective, clear demonstrations of how refinement-driven methodological choices translate into substantial and quantifiable reductions in animal use are still relatively uncommon. In the classic planning of our study, based on an independent-group comparison between infected and uninfected animals, sample-size calculations indicated the need for 26 females per group, totaling 52 animals to achieve adequate statistical power.

Guided by refinement principles, we redesigned the study using a paired approach, in which each female served as her own control, allowing direct comparison of reproductive outcomes before and after infection. This strategy reduced the required sample size to 14 females, while simultaneously decreasing interindividual variability and strengthening analytical robustness. In absolute terms, this methodological refinement resulted in a reduction of 38 animals, corresponding to a 73.1% decrease. Notably, this reduction was not achieved by compromising scientific rigor; rather, it emerged as a direct consequence of a refined design aligned with ethical and methodological best practices.

Taken together, this experience illustrates how refinement-oriented experimental planning can lead to concrete ethical gains while simultaneously improving data quality and interpretability. More importantly, we share this approach as a contribution to the broader scientific community, calling attention to the role of statistical reasoning and experimental design as practical tools for refinement. Beyond the specific context of genital leptospirosis, this strategy may be applicable to reproductive, pharmacological, toxicological, and other preclinical studies in which biologic variability is high and ethical considerations are paramount.

By communicating the line of reasoning that led from refinement to a substantial and quantifiable reduction in animal use, we aim to encourage similar methodological reflections and demonstrate that ethical responsibility and robust experimental design are not competing aims but complementary pillars of high-quality biomedical research.

Sincerely,

Isabel Roussoulieres, DVM, MS, FELASA Category B, Cristina Barbosa, DVM, PhD, and Walter Lilenbaum, DVM, PhD*

Department of Microbiology and Parasitology, Universidade Federal Fluminense, Niterói, Rio de Janeiro, Brazil

*Corresponding author. Email: wlilenbaum@id.uff.br

Conflict of Interest

The author declares that there are no competing interests.