Activity in Group-Housed Home Cages of Mice as a Novel Preclinical Biomarker in Oncology Studies

Abstract

Background: Improving experimental conditions in preclinical animal research is a major challenge, both scientifically and ethically. Automated digital ventilated cages (DVC^®) offer the advantage of continuous monitoring of animal activity in their home-cage. The potential utility of this technology remains understudied and deserves investigation in the field of oncology.

Methods: Using the DVC^® platform, we sought to determine if the continuous assessment of locomotor activity of mice in their home cages can serve as useful digital readout in the monitoring of animals treated with the reference oncology compounds cisplatin and cyclophosphamide. SCID mice of 14 weeks of age were housed in DVC^® cages in groups of four and followed with standard and digital examination before and after treatment over a 17-day total period.

Results: DVC^® detected statistically significant effects of cisplatin on the activity of mice in the short and long term, as well as trends for cyclophosphamide. The activity differences between the vehicle- and chemotherapy-treated groups were especially marked during the nighttime, a period when animals are most active and staff are generally not available for regular checks. Standard clinical parameters, such as body weight change and clinical assessment during the day, provided additional and complementary information.

Conclusion: The DVC^® technology enabled the home cage monitoring of mice and non-invasive detection of animal activity disturbances. It can easily be integrated into a multimodal monitoring approach to better capture the different effects of oncology drugs on anti-tumor efficacy, toxicity, and safety and improve translation to clinical studies.

Keywords: DVC®; SCID model; animal model; behavior; chemotherapeutic agents; cognitive impairment; digital biomarkers; drug development; home cage; oncology; preclinical; toxicity and safety; translation; welfare.

Figure 1

Design and body weight measurements. (A) A schematic representation of the study design. After 7 days of acclimation, mice in DVC^® systems were randomized into three study groups receiving the indicated agent or vehicle. Digital biomarkers (DVC^® locomotion index) and standard clinical information (body weight, appearance) were collected during the study. (B) Changes in body weight. Body weight was measured from Day −4 and every day for all three groups of mice. Relative weight changes from the baseline ± SD are shown. (two-way RM ANOVA; group p < 0.0001, day p < 0.0001, group × day p < 0.0001). The arrows indicate days of treatment.

Figure 2

Distribution of the clinical scores in the three study groups (n = 48) for (A) days corresponding to the baseline activity (Days −4 to −1) plus the day of the first treatment administration, Day 0; and (B) days following the first treatment administration. A two-way ANOVA-TYPE was applied to assess global effects with treatment group, day and group x day interaction; p < 0.0001, followed by a Bonferroni–Holm adjustment for each day and a Hochberg correction for each comparison to account for multiplicity across treatment and time. These statistics are summarized in Supplementary Table S1. Veh., vehicle; CPP, cyclophosphamide; Cis., cisplatin.

Figure 3

DVC^® locomotion index profile revealed alterations in animal activity upon chemotherapy intervention. Locomotion activity profiles generated from locomotion indexes in representative group-housed DVC^® cages from each treatment group: (A) vehicle, (B) cisplatin, (C) cyclophosphamide. Values were calculated from measurements collected over 1 h intervals. The gray areas correspond to the dark phase. The arrows indicate the days of injection.

Figure 4

DVC^® locomotion patterns confirmed alterations of animal activity upon chemotherapy intervention. (A) Heatmap of the dark phase from Days −4 to 13 for the three study groups. The heatmap shows activity data in the indicated groups with the locomotion index displayed in 12 h intervals (each row) during lights-off. Each column represents a DVC^® unit housing 4 mice. Colors denote the levels of calculated locomotion index, from blue (reduced locomotion index) to red (high locomotion index) (B) Heatmap of the light phase on Days −4 to 13 for the three study groups in 12 h intervals (C) heatmap depicting the full day activity (24 h intervals). The gray color indicates time slots for which no data were collected. Arrows indicate treatment intervention. Days of cage change “c” are specified.

Figure 5

Locomotion activity change revealed significant adverse effects of cisplatin on animal activity. Curves depicting the nocturnal and diurnal activity (night and day time activity). Locomotion activity change is presented as a percent of baseline ± SD, (A) at night (two-way ANOVA; study group (p < 0.0001), time (p = 0.0002), time by study group interaction (p = 0.0438)); (B) daytime (two-way ANOVA; study group (p = 0.1204); time (p < 0.0001), time by study group (p = 0.0317)); or (C) the entire day (two-way ANOVA; study group (p < 0.0001), time (p = 0.0002), time by study group interaction (p = 0.05)). Statistical comparisons for the different days are summarized in Supplementary Table S1.