Background: Rodent models are commonly used experimentally to assess treatment effectiveness in spinal fusion. Certain factors are associated with better fusion rates. The objectives of the present study were to report the protocols most frequently used, to evaluate factors known to positively influence fusion rate, and to identify new factors.

Method: A systematic literature search of PubMed and Web of Science found 139 experimental studies of posterolateral lumbar spinal fusion in rodent models. Data for level and location of fusion, animal strain, sex, weight and age, graft, decortication, fusion assessment and fusion and mortality rates were collected and analyzed.

Results: The standard murine model for spinal fusion was male Sprague Dawley rats of 295 g weight and 13 weeks’ age, using decortication, with L4-L5 as fusion level. The last two criteria were associated with significantly better fusion rates. On manual palpation, the overall mean fusion rate in rats was 58% and the autograft mean fusion rate was 61%. Most studies evaluated fusion as a binary on manual palpation, and only a few used CT and histology. Average mortality was 3.03% in rats and 1.56% in mice.

Conclusions: These results suggest using a rat model, younger than 10 weeks and weighing more than 300 grams on the day of surgery, to optimize fusion rates, with decortication before grafting and fusing the L4-L5 level.