Significance of Propionibacterium acnes-positive samples in spinal instrumentation

Abstract

Study design: A retrospective study about Propionibacterium acnes infections after Cotrel-Dubousset (CD) instrumentation.

Objectives: To analyze the significance of P. acnes-positive deep samples after CD.

Summary of background data: The diagnosis of spinal infections to P. acnes after CD is difficult.

Methods: Patients with revision surgery and at least 1 P. acnes-positive deep sample, between 2000 and 2006 were included. Group A had 1 revision surgery and group B had 2 successive revision surgeries, with P. acnes-positive deep samples. Group A was divided into 2 subgroups according to the peroperative macroscopic aspect, subgroup A1 with septic tissues, subgroup A2 without septic tissues. The biologic characteristics of the patients and the surgical and medical treatments were assessed.

Results: Sixty-eight patients were included, 60 in group A (A1 = 33, A2 = 27) and 8 in group B. Group A: 26 patients had 1 or 2 P. acnes-positive samples and 34 had at least 3 P. acnes-positive samples. Histology showed chronic inflammatory changes. C-reactive protein value median rate was 42 (A1) and 5 mg/L (A2). Twenty-two patients had a complete implant removal (14 with antibiotics, A1 = 12, A2 = 2). Nine patients had a total implant replacement (7 with antibiotics). Twenty-two patients had a partial implant removal (17 with antibiotics, A1 = 5, A2 = 12). Seven A1 patients had an irrigation and debridement (6 with antibiotics). The evolution was favorable for 28 patients. Seven patients had a documented sepsis. Group B: during the first revision, 8 patients had a partial implant removal (2 with antibiotics); during the second revision, all patients received antibiotics 4 of whom had a total implant removal. The long-term evolution was favorable for 6 patients.

Conclusion: P. acnes infection of spinal instrumentation is difficult to diagnose. Results of at least 4 deep sample cultures, histology, and C-reactive protein values must be compared to the peroperative macroscopic aspect.