Prognostic factors in spinal chordoma: A systematic review.

Objective: Discovering reliable prognostic factors for spinal chordoma remains a challenge. We attempted to identify evidence-based prognostic factors in the literature since its inception and to establish pooled relative risks (RR) of such factors.

Methods: MEDLINE and Embase search (inception to December 2014). Two reviewers independently selected papers involving spinal chordoma prognostic factors, and studied them for methodological quality and valuable new factors. Subsequently, we attempted to pool the results.

Results: Of 1465 citations, we studied 65 papers closely, and found several "new" prognostic factors. However, only eight papers were of adequate quality for analysis. Location in the upper cervical spine (pooled RR=5.46, 95% confidence interval [CI]: 2.23-13.34), worse preoperative Frankel score (pooled RR=2.77, 95% CI: 1.73-4.42), intralesional surgery (pooled RR=2.68, 95% CI: 1.66-4.32), greater extent of invasion (pooled RR=5.09, 95% CI: 1.49-17.41), and revision surgery (pooled RR=2.42, 95% CI: 1.34-4.36) appeared to be independent factors for worse outcome.

Conclusions: Despite the wealth of literature available, disappointingly few papers are of sufficient quality for drawing valid conclusions related to spinal chordoma prognostic factors. The heterogeneity of the studies renders results pooling almost impossible. More accurate individual prognostication requires methodologically high-quality studies with more uniform study design and data reporting.