Background & Aims

Bone fractures are one of the most common injuries worldwide. Currently, there is a paucity of studies that attempt to study post-fracture pain, therefore we set out to develop and validate a sensitive and unbiased methodology to measure pain following fracture in rodent models. We present our first data using a novel technology, the BlackBox System, coupled with DeepLabCut machine learning to monitor how quantifiable behavioral endpoints of pain shift after bone fracture. This technology allows for the collection of continuous data streams and automatically analyzes them to assess musculoskeletal pain behaviors rapidly and objectively in a natural setting.

Methods

In adult mice, we performed unstabilized tibia fractures and longitudinally assessed pain-related
behaviors and gait patterns at 4-, 11-, 18-, and 25-days post fracture (DPF). We used the BlackBox One system to perform high spatial and temporal recording of pain behaviors in freely moving mice, which captures both animal pose (body position) and paw weightbearing (paw pressure). We then performed automated extraction of hindpaw position and trained these algorithms to track changes in paw pressure and kinematic phenotypes using machine learning algorithms in DeepLabCut.

Results

First, we observe a significant decrease in weight bearing on the fractured limb as compared to the contralateral, non-fractured hindlimb in the fractured mice compared to the unfractured mice at day 4 that resolves around day 18. Secondly, we observe guarding (raised, folded paw) of the hindpaw of the fractured limb with peaks at days 4 & 11 in our fractured mice, a similarly well-characterized pain behavior in rodents. Next, we probed how pain-related changes in posture translate to changes in gait patterns during walking. To compensate for the loss of weight bearing on the injured hindlimb, we observed that fractured mice will often hop with their uninjured hindlimb.

Conclusions

Our data demonstrate how coupling the BlackBox system with powerful, high-throughput machine-learning tools (DeepLabCut) greatly enhances our ability to detect chronic pain behaviors in mice after fracture. This will allow for a more comprehensive analysis with greater sensitivity in quantifying induced and spontaneous pain-related behaviors than currently available methods (i.e., von Frey fibers, CatWalk/DigiGait, etc). By using the BlackBox, we will be able to assess the efficacy of non-opioid alternatives and therapeutics for post-fracture pain.

References

Basbaum AI, Bautista DM, Scherrer G, Julius D. Cellular and molecular mechanisms of pain. Cell. 2009 Oct 16;139(2):267-84. doi: 10.1016/j.cell.2009.09.028. PMID: 19837031; PMCID: PMC2852643.

Bahney CS, Zondervan RL, Allison P, Theologis A, Ashley JW, Ahn J, Miclau T, Marcucio RS, Hankenson KD. Cellular biology of fracture healing. J Orthop Res. 2019 Jan;37(1):35-50. doi: 10.1002/jor.24170. Epub 2018 Nov 30. PMID: 30370699; PMCID: PMC6542569.

Zhang Z, Roberson DP, Kotoda M, Boivin B, Bohnslav JP, González-Cano R, Yarmolinsky DA, Turnes BL, Wimalasena NK, Neufeld SQ, Barrett LB, Quintão NLM, Fattori V, Taub DG, Wiltschko AB, Andrews NA, Harvey CD, Datta SR, Woolf CJ. Automated preclinical detection of mechanical pain hypersensitivity and analgesia. Pain. 2022 Dec 1;163(12):2326-2336. doi: 10.1097/j.pain.0000000000002680. Epub 2022 May 11. PMID: 35543646; PMCID: PMC9649838.

Presenting Author

Jarret Weinrich

Poster Authors

Jarret Weinrich

PhD

University of California, San Francisco

Lead Author

E-mail

Allan Basbaum

Univ of California - San Francisco

Lead Author

E-mail

Chelsey Bahney

PhD

Steadman Phillippon Research Insitute

Lead Author

E-mail

Kazuhito Morioka

MD PhD

UCSF

Lead Author

E-mail

Charles Lam

UCSF

Lead Author

E-mail

Molly Czachor

Steadman Phillippon Research Insitute

Lead Author

E-mail

Topics

  • Models: Musculoskeletal