Introduction This abstract describes the development of an effective procedure for removing as much nucleus as possible from an intervertebral disc with minimal disruption to the annulus. The procedure was developed on cadaveric sheep discs which are well established as a model for human discs in studies of this kind. The purpose of the study was to develop a method for removing the nucleus as part of a laboratory study of nucleus replacement; however, it is also intended to guide the development of procedures for the removal of residual nucleus when indicated in surgical procedures that involve replacing the nucleus with synthetic materials.

Methods All procedures were performed via a 3 mm trocar. Four procedures were compared: (I) unilateral approach using rongeurs alone, (II) unilateral approach using rongeurs followed by chymopapain, (III) bilateral approach using rongeurs alone and (IV) bilateral approach using rongeurs followed by chymopapain. Chymopapain was administered as a solution (30 units in 0.1 cm³ de-ionised water) to a disc at 37°C. For each procedure (I–IV) 14 discs were used.

Results The percentages of nucleus removed were: (I) 34 ± 2%, (II) 41 ± 2%, (III) 52 ± 3% and (IV) 75 ± 8%; ANOVA showed a significant differences between the four sets of results (P < 0.05).

Conclusions Significantly more nucleus is removed using a bilateral than a unilateral approach; significantly more nucleus is removed if chymopapain is used in addition to rongeurs. A brush is useful in removing strands of nucleus loosened by chymopapain. For the purpose of these experiments a bottle brush with nylon bristles was trimmed to an overall diameter of about 10 mm, so that it could be inserted into the nuclear cavity via the trocar. Design of a surgical instrument for this purpose would be guided by a preliminary risk analysis.