Paraspinal muscles atrophy on both sides and at multiple levels after unilateral lumbar partial discectomy

Abstract

To identify the changes in cross-sectional areas (CSAs) and fatty infiltration of both sides of the paravertebral muscles and their associations with prognostic factors in patients who underwent unilateral lumbar discectomy. We retrospectively reviewed 27 patients who underwent magnetic resonance imaging before and after 1- or 2-level lumbar discectomy. The CSAs and functional cross-sectional areas of the paraspinal muscles were bilaterally measured from L1 to L2 to L5 to S1 based on T2-weighted axial images. These parameters were compared pre-and postoperatively. CSAs and functional cross-sectional areas decreased also in non-operative, non-surgical levels, not only in operated levels after discectomy. In the correlation analysis, the CSA of psoas major muscle at L1 to L2 was significantly decreased in patients with lower preoperative lordosis (r = 0.598, P = .040). The postoperative CSA of psoas major muscle at L4 to L5 was lower in those with the higher Pfirrmann grade (r = -0.590, P = .002); however, the CSA of quadratus lumborum muscle at L1 to L2 showed the opposite result (r = 0.526, P = .036). Similar results were also observed in the partial correlation adjusted for age and postoperative duration. Patients who underwent discectomy experienced overall paraspinal muscle atrophy in the lumbar region, including surgical and non-surgical sites. Such atrophic changes emphasized the need for core strengthening and lumbar rehabilitation from the early period after partial discectomy.

Figure 1.

Measurement of the cross-sectional area (CSA) of the paraspinal muscles on magnetic resonance imaging. (Left) Measurement of the total CSAs of the paraspinal muscles at L3–L4. (Right) Functional CSAs of the muscles measured using a threshold method, represented by the area-extracting highlighted part in green. CSA = cross-sectional area, m1, m2 = multifidus, m3, m4 = erector spinae, m5, m6 = psoas major, m7, m8 = quadratus lumborum.

Figure 2.

Flowchart of patient inclusion.

Figure 3.

(a), (b), and (c) pre- and postoperative comparisons of paraspinal muscle properties in the muscles of the operative side. The error bars represent ± 1 standard deviation of the group measurements. *P < .05, by paired t-test. ES = erector spinae, MF = multifidus, PM = psoas major, QL = quadratus lumborum.

Figure 4.

(a), (b), and (c) pre- and postoperative comparisons of paraspinal muscle properties in the muscles of the non-operative side. The error bars represent ± 1 standard deviation of the group measurements. *P < .05, by paired t-test. ES = erector spinae, MF = multifidus, PM = psoas major, QL = quadratus lumborum.

Figure 5.

The correlation between preoperative lordosis and Pfirrmann grade and relative changes in the CSA of the paraspinal muscles on the operated side. Relative changes in CSA were calculated by subtracting the pre-operative value from the post-operative value and dividing by the preoperative value. The pre-operative lumbar lordosis angle was measured as the angle formed by the extension line of the upper plate of the L1 vertebra body and the extension line of the lower plate of the L5 vertebra body. The preoperative lordosis and relative changes in the CSA of the PM at L1–L2 (A). The Pfirrmann scale grade and relative changes in CSA of the PM at L4–L5 (B) and the QL at L1–L2 (C). *P < .05, by the Pearson’s or Spearman`s correlation test. r, Pearson or Spearman correlation. CSA = cross-sectional area, ES = erector spinae, MF = multifidus, PM = psoas major, QL = quadratus lumborum.