Corrigendum

Stovell MG, Mada MO, Carpenter TA, et al. Phosphorus spectroscopy in acute TBI demonstrates metabolic changes that relate to outcome in the presence of normal structural MRI. J Cereb Bld Flow. Epub ahead of print 18 September 2018. DOI: 10.1177/0271678X18799176.

The authors wish to make the following corrections.

In the Results section of this article, in the section titled ³¹P MRS: metabolic changes in acute TBI, the sentence that reads

PCr/γATP was higher in TBI patients (Figure 2(a)): median (and interquartile range, IQR) PCr/γATP was 1.09 (1.00–1.22) in TBI patients and 0.91 (0.84–0.97) in healthy controls (p<0.0001).

has been corrected:

PCr/γATP was higher in TBI patients (Figure 2(a)): median (and interquartile range, IQR) PCr/γATP was 1.09 (1.04–1.20) in TBI patients and 0.93 (0.86–0.96) in healthy controls (p<0.0001).

The sentence that reads

TBI patients' brain pH was more alkaline than that of healthy controls (Figure 2(b)): median 7.03 (7.00–7.08) in TBI patients; and 7.00 (6.98–7.01) in healthy controls (p = 0.006).

has been corrected:

TBI patients' brain pH was more alkaline than that of healthy controls (Figure 2(b)): median 7.04 (7.02–7.05) in TBI patients; and 7.00 (6.99–7.00) in healthy controls (p<0.006).

The sentence that reads

This association was not strong (Spearman's correlation coefficient rho = 0.27), although highly statistically significant (p<0.0005).

has been corrected:

This association was not strong (Spearman's correlation coefficient rho = 0.27), although highly statistically significant (p<0.0001).

The sentence that reads

Grey matter/white matter ratio was not significantly different between TBI patients and healthy controls (p = 0.19, Mann-Whitney), nor did it reach significance as a covariate in linear mixed model analysis of PCr/γATP, PCr/total-mobilephosphate or pH.

has been corrected:

Grey matter/white matter ratio was not significantly different between TBI patients and healthy controls (p = 0.26, Mann-Whitney), nor did it reach significance as a covariate in linear mixed model analysis of PCr/γATP, PCr/total-mobilephosphate or pH.

The sentence that reads

However, it was a statistically significant covariate in linear mixed model analysis of γATP/total-mobile-phosphate (p = 0.0014), and as it can affect ratios of high-energy phosphates in the healthy brain,^14,22 it was included in all mixed effects model analysis.

has been corrected:

However, it was a statistically significant covariate in linear mixed model analysis of γATP/total-mobile-phosphate (p = 0.03), and as it can affect ratios of high-energy phosphates in the healthy brain,^14,22 it was included in all mixed effects model analysis.

In the Results section of this article, in the section titled ³¹P MRS: metabolic changes by outcome, the sentence that reads

Brain alkalosis (pH) was higher in patients with unfavourable outcome than patients with favourable outcome using both analytical methods (p<0.02) (Table 2).

Table 2.

Summary of ³¹P MRS results: group medians and interquartile ranges of within-subject means.

	Number subjects	Injured voxels	PCr/γATP Median (IQR)	PCr/total Median (IQR)	γATP/total Median (IQR)	pH Median (IQR)
TBI: all patients	13	Including	1.09 (1.04–1.20)	0.47 (0.46–0.48)	0.44 (0.40–0.45)	7.04 (7.02–7.05)
		Excluding	1.07 (1.06–1.21)	0.47 (0.46–0.49)	0.44 (0.40–0.45)	7.03 (7.01–7.05)
Healthy controls	10	NA	0.93 (0.86–0.96)	0.42 (0.41–0.43)	0.46 (0.45–0.48)	7.00 (6.99–7.00)
p (TBI vs. HC)		Including	<0.0001	<0.0001	0.0009	0.042
		Excluding	<0.0001	<0.0001	0.001	0.055
TBI: favourable outcome	7	Including	1.07 (1.04–1.10)	0.48 (0.46–0.48)	0.44 (0.44–0.45)	7.02 (7.00–7.03)
		Excluding	1.07 (1.06–1.09)	0.47 (0.46–0.48)	0.44 (0.44–0.45)	7.02 (7.00–7.03)
TBI: unfavourable outcome	6	Including	1.14 (1.07–1.22)	0.47 (0.46–0.49)	0.41 (0.39–0.42)	7.07 (7.04–7.14)
		Excluding	1.14 (1.06–1.24)	0.48 (0.47–0.49)	0.41 (0.40–0.44)	7.07 (7.03–7.16)
p (Fav vs. Unfav.)		Including	0.3	0.9	0.04	0.03
		Excluding	0.3	0.9	0.06 [0.04]	0.057 [0.008]

Note: ³¹P magnetic resonance spectroscopy (MRS) measurements of pH, phosphocreatine (PCr), adenosine triphosphate (γATP) and ‘total' mobile phosphate (PCr + γATP + inorganic phosphorus) ratios in 13 patients suffering from acute major traumatic brain injury (TBI) and 10 age-group matched healthy controls (HC). Results represent group medians of within-subject means, with interquartile ranges (IQR) in parentheses (curved brackets); for each individual subject's mean values see Table 3. Voxels were considered injured if they contained ≥5% radiological injury on fluid-attenuated inversion recovery (FLAIR) or susceptibility weighted injury (SWI) MR sequences. Favourable outcome (Fav) defined as six-month Glasgow outcome score (extended) (GOS-E) ≥4, and unfavourable outcome (Unfav.) GOS-E ≤ 3. Statistical analysis performed using a linear mixed model in R (lme in package nlme), adjusting for voxel grey matter/white ratio. Inclusion of healthy controls and analysis of the linear model using glht (results in Figure 4) found the difference in γATP/total-mobile-phosphate to be statistically significant (results in square [ ] brackets), but did not affect the significance of other measurements.

has been corrected:

Brain alkalosis (pH) was higher in patients with unfavourable outcome than patients with favourable outcome using both analytical methods (p<0.03) (Table 2).

In the Results section of this article, in the section titled ³¹P MRS: metabolic changes in the absence of visible tissue injury, the sentence that reads

The difference in γATP/total-mobile-phosphate narrowly lost significance, both by lme (p = 0.076) and by Mann-Whitney (p = 0.18).

has been corrected:

The difference in γATP/total-mobile-phosphate narrowly lost significance, both by lme (p = 0.06) and by Mann-Whitney (p = 0.18).

In Figure 4, there is a change to the second p value in part (a). It previously read p = 0.37 and has been corrected as below, p = 0.28. Figure 4.

In Table 2, the value of p (Fav vs Unfav.) in the column PCr/total Median (IQR) and the value of p (TBI vs. HC) in the column pH Median (IQR) have been corrected as below.

In Table 3, the values of H01-H10 in column PCr/γATP should have appeared as below.

Table 3.

³¹P MRS data acquired from TBI patients and healthy controls.

Subject ID	PCr/γATP mean	PCr/γATP S.D.	PCr/total mean	PCr/total S.D.	γATP/total mean	γATP/total S.D.	pH mean	pH S.D.	No. of voxels
P01	1.06	0.11	0.47	0.03	0.45	0.02	7.03	0.09	7
P02	1.10	0.17	0.48	0.04	0.44	0.03	7.02	0.03	7
P03	1.41	0.13	0.52	0.04	0.37	0.03	7.29	0.07	7
P04	1.00	0.12	0.42	0.06	0.42	0.04	7.15	0.09	7
P05	1.09	0.23	0.46	0.06	0.42	0.04	7.10	0.07	4
P06	1.07	0.14	0.48	0.04	0.45	0.02	7.01	0.07	6
P07	1.03	0.13	0.45	0.03	0.45	0.02	6.98	0.02	8
P08	1.21	0.13	0.48	0.04	0.40	0.04	7.02	0.06	8
P09	1.20	0.04	0.47	0.02	0.39	0.02	7.05	0.05	6
P10	1.23	0.09	0.49	0.03	0.40	0.03	7.04	0.02	7
P11	1.04	0.05	0.46	0.01	0.44	0.02	7.04	0.02	7
P12	1.04	0.08	0.46	0.03	0.44	0.01	7.04	0.07	8
P13	1.10	0.28	0.50	0.06	0.46	0.05	6.96	0.06	8
H01	0.95	0.07	0.42	0.02	0.45	0.01	6.99	0.02	8
H02	0.74	0.07	0.37	0.02	0.50	0.02	7.00	0.02	8
H03	0.92	0.04	0.42	0.03	0.46	0.03	7.00	0.04	8
H04	0.98	0.05	0.44	0.02	0.45	0.01	7.00	0.01	8
H05	0.82	0.05	0.40	0.02	0.49	0.03	6.99	0.02	8
H06	0.99	0.04	0.45	0.01	0.45	0.01	6.97	0.03	8
H07	0.95	0.06	0.43	0.02	0.45	0.01	7.00	0.01	8
H08	0.88	0.04	0.42	0.02	0.48	0.01	6.99	0.04	8
H09	0.86	0.04	0.41	0.01	0.48	0.01	7.00	0.03	8
H10	0.97	0.04	0.43	0.03	0.45	0.03	7.03	0.03	8

Note: Individual subject ³¹P magnetic resonance spectroscopy (MRS) measurements of pH, phosphocreatine (PCr), adenosine triphosphate ( γ ATP) and ‘total’ mobile phosphate (PCr +  γ ATP + inorganic phosphorus (Pi)) ratios derived from the central eight voxels of 13 TBI patients (P1-P13) and 10 age-group-matched healthy controls (H1-H10). As part of a separate biochemical study, some patients had microdialysis catheters supplemented with either succinate or glucose so a voxel was excluded from analysis (P01, P02, P03, P04, P05, P06, P10, P11). Two subjects’ CSI grids were positioned to capture superficially placed microdialysis catheters, excluding a further one (P06 and P09) and three (P05) voxels that represented < 90% brain tissue.