Multifidus muscle changes and clinical effects of one-level posterior lumbar interbody fusion: minimally invasive procedure versus conventional open approach

Abstract

We set out to determine whether a minimally invasive approach for one-level instrumented posterior lumbar interbody fusion reduced undesirable changes in the multifidus muscle, compared to a conventional open approach. We also investigated associations between muscle injury during surgery (creatinine kinase levels), clinical outcome and changes in the multifidus at follow-up. We studied 59 patients treated by one team of surgeons at a single institution (minimally invasive approach in 28 and conventional open approach in 31, voluntarily chosen by patients). More than 1 year postoperatively, all the patients were followed up with the visual analogue scale (VAS) and Oswestry disability index (ODI), and 16 patients from each group were evaluated using MRI. This enabled the cross-sectional area (CSA) of lean multifidus muscle, and the T2 signal intensity ratio of multifidus to psoas muscle, to be compared at the operative and adjacent levels. The minimally invasive group had less postoperative back pain (P < 0.001) and lower postoperative ODI scores (P = 0.001). Multifidus atrophy was less in the minimally invasive group (P < 0.001), with mean reductions in CSA of 12.2% at the operative and 8.5% at the adjacent levels, compared to 36.8% and 29.3% in the conventional open group. The increase in the multifidus:psoas T2 signal intensity ratio was similarly less marked in the minimally invasive group where values increased by 10.6% at the operative and 8.3% at the adjacent levels, compared to 34.4 and 22.7% in the conventional open group (P < 0.001). These changes in multifidus CSA and T2 signal intensity ratio were significantly correlated with postoperative creatinine kinase levels, VAS scores and ODI scores (P < 0.01). The minimally invasive approach caused less change in multifidus, less postoperative back pain and functional disability than conventional open approach. Muscle damage during surgery was significantly correlated with long-term multifidus muscle atrophy and fatty infiltration. Furthermore these degenerative changes of multifidus were also significantly correlated with long-term clinical outcome.

Fig. 1

Measurement of cross-sectional area of multifidus muscle in a highly atrophied muscle (left). Lean muscle CSA was the sum of all the areas outlined in yellow. The ROI outlined for T2 signal intensity of gross multifidus muscle is shown on the right

Fig. 2

Serum creatinine kinase (CK) concentrations for both groups. CK levels were significantly lower in the MI group than the CO group at 1, 3 and 5 days after surgery. CK concentrations of both groups returned to baseline after 7 days

Fig. 3

Scatterplots showing associations among CK Level, VAS and ODI at the final follow-up, and changes of multifidus CSA and T2 signal intensity ratio of multifidus to psoas muscle. CK level versus CSA (a) and intensity ratio (b); VAS score versus CSA (c) and intensity ratio (d); and ODI score versus CSA (e) and intensity ratio (f). All the associations were significant