Should We Cross the Cross-links?

• Published in: Spine (Philadelphia, Pa. 1976) Pa. 1976), 2013-08, Vol.38 (18), p.E1128-E1134
• Main Authors: Kulkarni, Arvind G., Dhruv, Abhilash N., Bassi, Anupreet J.
• Format: Article
• Language: English
• Subjects: Biomechanical Phenomena - physiology; Follow-Up Studies; Humans; Prostheses and Implants - utilization; Radiography; Retrospective Studies; Spinal Diseases - diagnostic imaging; Spinal Diseases - surgery; Spinal Fusion - instrumentation; Spinal Fusion - methods
• ISSN: 0362-2436; 1528-1159
• DOI: 10.1097/BRS.0b013e31829af99f

STUDY DESIGN.Retrospective study. OBJECTIVE.To assess critically if cross-links are necessary adjuvants in posterior spinal constructs. SUMMARY OF BACKGROUND DATA.Although numerous biomechanical studies are available in the literature, there has been no clinical study that has evaluated the need for cross-links in clinical situations. METHODS.The spinal constructs of patients of varied etiology who underwent surgery between July 2007 and July 2011 without the usage of cross-links were evaluated. The immediate postoperative erect radiographs were compared with the erect radiographs at the last follow-up by 2 independent observers (spine fellows not involved in the management of the patients) critically for any rotational instability using the Nash-Moe technique of assessment of vertebral rotation as well as for any “parallelogram effect.” The intraobserver and interobserver reliability was analyzed. RESULTS.There were 208 cases included in the study during the study period that satisfied the criteria. The total number of motion segments fused was 707 ranging from 1 to 15 involving various etiologies. The average follow-up was 15 months (12–36 mo). Barring one patient with a thoracolumbar fracture with rotational instability (AO [Arbeitsgemeinschaft für Osteosynthesefragen] type C) who had undergone a short-segment fixation, none of the cases demonstrated any rotational instability in the follow-up radiographs. Interestingly, the rotational instability (parallelogram effect) in that patient got corrected spontaneously once anterior reconstruction was performed. The intraobserver reliability was 100% and the interobserver reliability was 92.83%. This variability was in assessing the grade of vertebral rotation only; none of the levels had a change in rotation irrespective of variation in grade assessment in the final postoperative radiograph. CONCLUSION.This study concludes that use of cross-links in clinical practice may be avoidable. The derivations from biomechanical studies do not translate into clinical advantages. Eliminating the usage of cross-links reduces the operative time as well as the overall total hospital costs (a single cross-link may cost anywhere between $1500 and $2000 and surgeons tend to use single or multiple cross-links). Additionally, prominence of implants, corrosion, infection, implant failure, and pseudarthrosis are the other complications attributed to cross-links in the literature that can be eliminated by preventing their incorpo