Reduced temporal muscle thickness predicts shorter survival in patients undergoing chronic subdural haematoma drainage

Abstract

Background: Chronic subdural haematoma (CSDH) drainage is a common neurosurgical procedure. CSDHs cause excess mortality, which is exacerbated by frailty. Sarcopenia contributes to frailty - its key component, low muscle mass, can be assessed using cross-sectional imaging. We aimed to examine the prognostic role of temporal muscle thickness (TMT) measured from preoperative computed tomography head scans among patients undergoing surgical CSDH drainage.

Methods: We retrospectively identified all patients who underwent CSDH drainage within 1 year of February 2019. We measured their mean TMT from preoperative computed tomography scans, tested the reliability of these measurements, and evaluated their prognostic value for postoperative survival.

Results: One hundred and eighty-eight (122, 65% males) patients (median age 78 years, IQR 70-85 years) were included. Thirty-four (18%) patients died within 2 years, and 51 (27%) died at a median follow-up of 39 months (IQR 34-42 months). Intra- and inter-observer reliability of TMT measurements was good-to-excellent (ICC 0.85-0.97, P < 0.05). TMT decreased with age (Pearson's r = -0.38, P < 0.001). Females had lower TMT than males (P < 0.001). The optimal TMT cut-off values for predicting two-year survival were 4.475 mm for males and 3.125 mm for females. TMT below these cut-offs was associated with shorter survival in both univariate (HR 3.24, 95% CI 1.85-5.67) and multivariate (HR 1.86, 95% CI 1.02-3.36) analyses adjusted for age, ASA grade and bleed size. The effect of TMT on mortality was not mediated by age.

Conclusions: In patients with CSDH, TMT measurements from preoperative imaging were reliable and contained prognostic information supplemental to previously known predictors of poor outcomes.

Keywords: Body composition; Chronic subdural haematoma; Computed tomography; Frailty; Sarcopenia.

Figure 1

Study flowchart.

Figure 2

(A) The semi‐axial plane used for the temporal muscle measurements. The yellow line connects the most medial part of the bony temporal fossa and the most lateral part of the temporoparietal bone without intersecting the cranium. The temporal muscle measurements (red lines, white arrowheads) were made perpendicular to this line on the thickest part of the temporal muscle. (B–D) The axial, coronal and sagittal localisation planes used to align the scan to produce the sem‐iaxial plane A. The yellow lines are reference lines connecting the insertions of the falx cerebri (B), parallel to the middle fossa floor (C), and the anterior skull base (D).

Figure 3

(A) Age‐related distribution of mean temporal muscle thickness among 188 patients with surgically managed chronic subdural haematoma. A locally weighted scatterplot smoothing curve (unbroken line) is shown. The dashed lines represent the upper limits of the first, second, and third age quartiles from left to right (at 70.0, 78.4, and 85.3 years, respectively). Temporal muscle thickness was inversely correlated with age at operation. (B) Boxplot of mean temporal muscle thicknesses of each age quartile. Only between‐group comparisons with P < 0.05 are shown. (C) Boxplot showing that male patients had thicker temporal muscles than female patients.

Figure 4

The receiver operating characteristic (ROC) curves for temporal muscle thickness (TMT) used to predict 2‐year mortality. The diagonal lines are reference lines demonstrating equal sensitivity and specificity, and the intersection of the dashed lines represents the highest Youden's index calculated from the ROC curve. The resulting TMT cut‐off values were 4.475 mm for males and 3.125 mm for females. AUC, area under the curve.

Figure 5

Cox regression survival curves for overall survival following chronic subdural haematoma surgery stratified by the sex‐specific cut‐off values of temporal muscle thickness. Variables included in the model were TMT according to the sex‐specific cut‐off values, age at operation, American Society Anesthesiologists Physical Classification Scale score and haematoma volume (Table 3, Model 2).