INTRODUCTION: Lumbar discography has been controversial for many years. Some of the controversy has been based on concerns about the needle causing harm to the injected, non-operated discs. However, other studies suggest discography does not damage discs. Long-term follow-up of injected discs may be helpful to determine if discography was related to development of clinically relevant disc degeneration, evidenced by subsequent surgery. The purpose of this study was to compare rates of disc-related surgery during long-term follow-up in discs that received discography vs. those that did not.
METHODS: Patients were identified from a large-scale retrospective study evaluating lumbar total disc replacement (TDR) outcomes. Data collected included patient descriptives, discography information, and details of re-operations. For patients who had not been seen recently in the clinic, mailings and telephone calls were conducted attempting to collect current follow-up data, including re-operation. Only patients for whom minimum 10-year follow-up data were available were included to better assess the potential development of disc degeneration post-discography. The mean follow-up was 156.1 months (range 120 to 254 months). Data were analyzed by individual disc level. A level was included if it was: L3-4, L4-5, L5-S1, or L2-3 that underwent discography or was adjacent to an operated disc. Discs were excluded if they underwent TDR as the index surgery or had previously undergone TDR or interbody fusion. The study included 251 discs that received discographic injection and 124 discs that did not. Rates of surgery for disc degeneration or herniation after the index surgery for these two groups were compared.
RESULTS: The rate of surgery for disc-related pain was not greater among discs that underwent discography vs. those that were not injected (10.8% vs. 8.1% p>0.40). Forward conditional binary logistic regression analysis found that the factor most strongly related to subsequent disc-related surgery was the individual lumbar level. Once in the equation the only other variable significantly related to re-operation was the result of the discogram (not simply having a discogram itself). Age, body mass index, sex, or having a discographic injection were not significantly related to subsequent surgery. Based on the results of the regression analysis, relationships between subsequent surgery and level and discography results were investigated further. Re-operation for disc-related pain was most common at L5-1 (19.6%) and L4-5 (13.6%), regardless of discography performed or not. Rates of surgery for disc-related pathology were similar for discs not receiving discography (8.1%) and those that were normal on discography (6.3%). These were both significantly less than the rate of 18.4% among discs that were not normal on discography (painful upon injection and/or discordant pain, and/or abnormal in appearance on imaging).
DISCUSSION: The results of this study found there was no significant difference in the rate of disc-related surgery among discs that underwent discography vs. those that did not, at a minimum follow-up of 10 years. The rate was greater only among discs that were abnormal with discography. These findings support that discographic injection does not injure normal discs.