Skip to main content
Account

Table 3 The use of the callosal angle on MRI and CT for prediction of shunt response in iNPH

From: Radiological predictors of shunt response in the diagnosis and treatment of idiopathic normal pressure hydrocephalus: a systematic review and meta-analysis

Study

Sample size

Radiological methodology

Cutoff specification

Image specification

Image plane

Main reported outcomes

Mantovani et al. [33]

n = 62

•Callosal angle: level of the posterior commissure orthogonal to the anterior commissure-posterior commissure (ACPC) line

•Anterior callosal angle: at the level of the anterior commissure orthogonal to ACPC line

•CA: 59.5°

•ACA: 112°

•3 T MRI

•Coronal plane

•Mean ACA was higher than mean CA, 103.6° ± 14.2° vs 58.9° ± 16.1° (p < 0.001)

•Neither CA or ACA were found to significantly predict mRs or INPHGS outcome

ACA:

•There was a negative correlation between ACA and Tinetti total score (r = -0.306, p < 0.05)

•Mean ACA in SR patients was smaller than SNR patients, (98.3° ± 11.4° and 108.6° ± 15.1° respectively)

•Mean ACA was significantly smaller in those with reduced fall risk post shunt

•Using a cutoff has a Youden’s Index = 0.344. The OR for ACA between SR and SNR is 2.97 (95% CI 1.04–8.5),

CA:

•There were no significant differences between SR. Values not given

•There was no effect of CA size on fall risk post shunt

•Using a cutoff shows Youden’s Index = 0.327. The OR between SR and SNR is 2.15 (95% CI 1.03–4.52)

Virhammar et al., 2014 [60]

n = 108

•Angle between lateral ventricles through the posterior commissure

•N/A

•MRI. T2 Flair, T1-weighted MRI. (9% of patients on 3 T scanner; 70% on a 1.5 T scanner, 14% on a 1 T scanner and 7% on a 0.5 T scanner

•Coronal plane, image taken perpendicular to the anterior/ posterior commissure plane

OR between SR and SNR: 0.57[(0.36–0.91) p = 0.017]

Virhammar et al., 2014 [59]

Callosal angle only

n = 108

•Angle between the lateral ventricles through the posterior commissure

•63°

•MRI 3D T1-weighted images. Ten (9%) on 3-T, 75 (69%) on 1.5-T, 16 (15%) on 1-T, and 8 (7%) on 0.5-T

•Coronal plane through posterior commissure, perpendicular to the anterior commissure -Posterior commissure plane

•CA was significantly smaller in SR [59° (95% CI 56°–63°)] than SNR [68° (

•95% CI 61°–75°)] (p < 0.05)

•Multivariate analysis: Smaller CA was significantly associated SR (OR 0.97, [95% CI 0.94–0.99], p < 0.05)

Cutoff: Sensitivity: 0.67, Specificity:0.65, Youden’s index: 0.33

•TP: 55, TN: 18, FP: 9, FN:27. N = 109

•Weak inverse correlation between EI and CA (r = -0.23, p < 0.05)

Narita et al. [41]

n = 103

•Angle between the left and right corpus callosum

•N/A

•3D T1-weighted MRI obtained with a Signa 1.5 T MR imaging unit

•Coronal plane at the posterior commissure

•Simple linear regression analysis showed significant association between CA and MMSE improvement (B =  − 0.04, R2 = 0.08, p = .035). There was no significant association between CA and Total score, gait cognitive or urinary subjections or TUG (B =  − 0.02, − 0.01, − 0.01, − 0.01, − 0.12 respectively)

Hong et al. [20]

n = 31

•Angle between the lateral ventricles

•N/A

•3.0 Tesla MRI scanner was used to gain Axial fluid-attenuated inversion recovery (FLAIR), T2- weighted images, T1-weighted images, and coronal T1-weighted images

•Coronal plane through posterior commissure, perpendicular to the anterior commissure -Posterior commissure plane

•Difference in mean CA in SR: 75.2 ± 21.7 and in SNR groups: 88.3 ± 18.2 was not significant (p = 0.109). No other statistical analysis was performed

Agerskov et al. [2]

n = 168

•N/A

•N/A

•MRI 1.5 T. T1-weighted images

•Coronal plane at level of the posterior commissure

•All patients had CA < 90°

•There was no difference in CA findings between SR (median 68°) and SNR (median 69°) [p > 0.05] and it could not be used to predict SR in multivariate logistical analysis

•Its non-significant correlation coefficient with the composite score was 0.17

Grahnke et al. [16]

n = 72

•Angle at the level of midpoint of corpus collosum

•105.4°

•CT or MRI

•Mid-sagittal plane, parallel to floor of 4th ventricle

•Mean CA in SR was 108.4 (SD: 16.8) while SNR was 117.6 (SD: 14.2), p = 0.030. Diagnostic accuracy AUC of 0.64 95% CI (0.50–0.78). Cutoff of 105.4 has sensitivity 0.415 and a specificity of 0.87. A patient is 4% more likely to have post-shunt benefit for every degree CA is lower: OR (unadjusted) 0.96 [(95% CI:0.93–0.998) p = 0.037], adjusted OR: 0.96 [(95% CI: 0.93–0.997) p = 0.036]

•TP: 19, TN:23, FP:4, FN:26

Black [5]

n = 62

•Angle of the junction of frontal horn roofs

•120°

•Pneumoencephalogram

•AP projection

•There was no significant difference in CA between SR and SNR. The cutoff had a sensitivity of 50%, specificity of 60%, PPV of 42.9% and NPV of 66.7%. TP:3, FP:4, TN:6, FN:3

  1. Studies included assessing the use of any MRI or CT callosal angle as predictor of shunt responsiveness. MRI studies are above the double solid lines, CT studies are below. SR, shunt response; S-NR, shunt non-response; ACA, anterior callosal angle; CA, callosal angle; mRS, modified Rankin scale; MMSE, mini mental state examination; TUG, timed up and go test; OR, odds ratio; CI, Confidence interval; AUC, area under the curve; SD, standard deviation; NPV, negative predictive value; PPV, positive predictive value; TP, true positives; FP, false positives; TG, true negatives; FN, false negatives