Sage Journals: Discover world-class research

Abstract

The purposes of this study are to assess the oxygen extraction fraction (OEF) changes on MRI-based quantitative susceptibility mapping (QSM) in systemic lupus erythematosus (SLE) patients and to determine whether QSM-OEF is associated with disease activity in SLE. We enrolled 42 SLE patients and 20 healthy subjects (HS) who had no pathologies on conventional brain MRI. Disease activity was assessed using SLE Disease Activity Index (SLEDAI). For the measurement of QSM-OEF, QSM data were analysed using the Perfusion Mismatch Analyzer software program. Spearman's or Pearson's correlation coefficients were calculated, and independent predictors were identified through a multiple linear regression analysis. QSM-OEF was significantly higher in SLE than that in HS (51.3 ± 10.1 vs. 40.5 ± 3.7, p < 0.001). QSM-OEF was positively correlated with SLEDAI and the presence of neuropsychiatric symptom (NPS) scores (ρ = 0.663, p < 0.001 and ρ = 0.340, p = 0.028). At multiple linear regression analysis, SLEDAI and NPS were independently associated with QSM-OEF (standardized β = 0.426, p = 0.016 and standardized β = 6.148, p = 0.029). In the SLE patients, QSM-OEF is associated with disease activity, which might predict an increased risk of stroke in SLE.

Keywords

Systemic lupus erythematosus oxygen extraction fraction quantitative susceptibility mapping perfusion mismatch analyzer misery perfusion

Introduction

Systemic lupus erythematosus (SLE) is an autoimmune disease that predominantly affects the central nervous system (CNS) in women of child-bearing age.¹ The most prevalent manifestation of lupus is neuropsychiatric systemic lupus erythematosus (NPSLE), occurring in as many as 30%–56% of all SLE patients.^2,3 NPSLE patients have higher morbidity and mortality rates than SLE patients with non-neuropsychiatric syndrome (non-NPSLE).⁴ Large infarcts, cortical atrophy and multifocal gray matter (GM) and/or white matter (WM) lesions are frequently observed on magnetic resonance imaging (MRI) in NPSLE patients.⁵ However, a substantial population of patients with NPSLE exhibit normal MRI findings and present with only neuropsychiatric symptoms (NPS),⁶ as the metabolic and/or functional alterations of the disease usually may not have anatomic disturbances detectable on conventional MRI. Therefore, methods other than conventional MRI are needed to evaluate brain changes in NPSLE.

Oxygen extraction fraction (OEF) represents an important measurable parameter of brain metabolism,⁷ which provides information about the relative deficiencies in cerebral blood supply with the tissue's oxygen demand, so-called ‘misery perfusion’.⁸ The risk of stroke recurrence increases with misery perfusion in patients with symptomatic major cerebral arterial occlusive disease.^9,10 Brain pathophysiology studies in NPSLE patients have shown that vasculitis in the intramural blood vessels of arterioles, capillaries and venules¹¹ causes brain parenchymal lesions secondarily.¹² Thus, we hypothesised that the disease activity related to vasculitis and hypoxia may increase the OEF in NPSLE patients. OEF, therefore, would seemingly provide additional information, especially in the monitoring phase of the disease severity.

Traditional approaches with ¹⁵O positron emission tomography (PET) are generally considered to be the gold standard approach for OEF measurements; however, there are some notable drawbacks with the technique, such as the limited availability of single-photon emission computed tomography (SPECT) scanners and considerable radiation exposure, which may prevent the potential for repeat scans, especially for young patients including SLE, and the extended scan time that is required. Furthermore, ¹⁵O PET cannot be readily applied in the clinics because of the complex setup necessary to administer short half-life radiotracers. Recent advancements in the field of MRI have ushered several new approaches to OEF measurement using quantitative susceptibility mapping (QSM) of MRI. The QSM technique is insensitive to the geometry of veins and dipolar artefacts; both of which affect the measured distribution of measured phase values. Kudo et al.¹³ reported that a good correlation was obtained between OEF measurements using QSM of MRI and the gold standard ¹⁵O PET in the identification of elevated OEF in affected hemispheres of patients with unilateral chronic steno-occlusive disease. The purpose of this study was to assess the QSM-OEF changes in SLE patients and to determine whether QSM-OEF is associated with disease activity in SLE.

Material and methods

Approval

Our study was approved by the Institutional Review Board of the University of Occupational and Environmental Health School of Medicine (Kitakyushu, Fukuoka, Japan), which waived informed consent because it was retrospective.

Subjects

One radiologist with seven years of neuroradiology experience (M.M.) reviewed the MRI database of patients who were examined between May 2004 and December 2016 who had been diagnosed with SLE according to the American Rheumatism Association criteria for the classification of SLE¹⁴ and selected 60 patients. The disease activity of SLE patients was quantified using a Safety of Estrogens in Lupus Erythematosus National Assessment-SLE Disease Activity Index (SELENA-SLEDAI) score¹⁵ and the British Isles Lupus Assessment Group 2004 (BILAG) index score.¹⁶ We further selected 48 of the 60 patients in whom disease activity was assessed within 15 days of their MRI examinations using the SLEDAI and BILAG index scores. The exclusion criteria included patients with a history of other neurological diseases, unsatisfactory images due to artifacts, and/or MRI abnormalities. With regard to abnormal MR findings, patients with mild or moderate brain atrophy (see ‘Statistical analysis’) and mild WM lesion (WML) (Fazekas grade 1) on T2-weighted imaging (T2WI)/fluid-attenuated inversion recovery (FLAIR) imaging were not excluded.¹⁷ Of the six excluded patients, five had a history of major infarction and one had a motion artifact.

Thus, we enrolled 42 SLE patients (38 women, 4 men; mean age 38.7 ± 15.8 years, range 19–66 years) and 20 healthy subjects (HS) (19 women, 1 man; mean age 40.7 ± 13.0 years, range 22–62 years) to QSM and conventional T2WI/FLAIR imaging. HS were age- and sex-matched with the SLE group, had no history of neurological or psychiatric diseases and underwent MRI including QSM between January 2008 and December 2016. The indications for their examination included screening, headaches or benign positional vertigo.

We reviewed the duration of SLE (i.e. the interval between the initial onset of SLE and the brain MRI study) and demographic data for vascular risk factors (diabetes mellitus defined as a random glucose level of > 11.1 mmol/l, fasting blood glucose > 7.0 mmol/l, hemoglobin (Hb) A1c > 6.5% or the current use of antidiabetic drugs); hypertension (blood pressure > 140/90 mm Hg or current treatment with anti-hypertensive drugs); past or current smoking; dyslipidemia (low-density lipoprotein cholesterol > 3.64 mmol/l, high-density lipoprotein cholesterol < 0.91 mmol/l, triglyceride > 1.7 mmol/l or treatment for dyslipidemia); obesity (body mass index > 26 kg/m²); and treatments (a history of pulse steroid therapy).

MRI

All subjects underwent brain MRI with QSM. All studies were performed on a 3T MR imaging system (Signa Excite 3T; GE Health Care, Milwaukee, Wisconsin, US) using a dedicated 8-channel phased array coil (Invivo, Gainesville, Florida, US). QSM was obtained with a 3D flow-compensated multi-echo spoiled gradient echosequence.¹⁸ The imaging parameters included the following: coronal planes covering the brain; number of TEs: 11; first TE: 4.5 ms; TE spacing: 5 ms; TR: 58.4 ms; flip angle: 15 °; bandwidth per pixel: ± 62.5Hz; FOV: 22 × 16.5 cm; acquisition matrices: 320 × 416; section thickness: 2 mm; imaging time: 7 min, 1 s. A parallel imaging method (the array spatial sensitivity encoding technique) was used with a reduction factor of 2. In this study, 56 axial sections were used to capture the centre of the brain that contains the deep grey nuclei structures. QSM was reconstructed from the complex data obtained during the gradient-echo sequence by using the morphology-enabled dipole inversion technique.¹⁹ After the complex multi-echo MR images were saved, a nonlinear fitting was performed to estimate the magnetic field inhomogeneity, followed by a magnitude-guided phase unwrapping.²⁰ The background field was further removed by applying the projection to the dipole fields method.²¹ Finally, the remaining tissue field was inverted to generate a susceptibility map by using the morphology-enabled dipole inversion method.²⁰ An empirically determined regularisation parameter of 1000 was consistently applied to all cases. All 11 echoes were used with weightings proportional to their signal-to-noise ratios for reconstructing QSM.^22,23

Data processing of OEF

Co-registration of QSM-OEF images was performed using MRI magnitude images and SPM8 software (http://www.fil.ion.ucl.ac.uk/spm/). The registered QSM-OEF images were then imported into a Perfusion Mismatch Analyzer software (version 4.0.4),²⁴ where automated measurement of regions-of-interests (ROIs) was performed using a previously described method (Figure 1).^24,25

Figure 1.

Creation of OEF map. QSM in 31-year-old female with severe SLE disease activity (SLEDAI score = 17, BILAG index score = 22) is shown (a). Bright veins indicate amount of deoxygenated hemoglobin. Venous mask image (b) is obtained by using local threshold method. OEF map (c) is created from susceptibility value of these veins. (d) ROI measurements of QSM-OEF. Example of automated measurement of ROIs is shown. Eighteen spherical ROIs are automatically placed along the entire parenchymal surface of the brain. Blue line indicates midline of the brain.

An estimate of QSM-OEF (1-oxygen saturation) can be obtained using following equation OEF =A × hematocrit (Hct) × 1/B.

A is susceptibility difference between the average susceptibility of veins and surrounding tissue was calculated in each ROI. B (1.8 × 10⁻⁷ in CGS units) is the susceptibility difference between fully deoxygenated blood and fully oxygenated blood. For SLE patients, we used actual measurement data of Hct because of high frequency of SLE patients with anaemia. For healthy subjects, we assumed Hct value of 0.40 for female and 0.45 for male.

On a slice at the level of centrum semiovale, 18 spherical ROIs with a diameter of 10 mm (20 pixels) were automatically placed along the entire parenchymal surface of the brain without overlapping each other (Figure 2). For all HS and SLE patients, the mean OEF value of both hemispheres (nine ROIs for each hemisphere) was calculated.

Figure 2.

QSM (a) and OEF map (b) in Case A (a 31-year-old woman) with mild SLE activity (SLEDAI score = 0, BILAG index score = 4). QSM (c) and OEF map (d) in Case B (a 36-year-old woman) with severe SLE activity (SLEDAI score = 25, BILAG index score = 22). QSM-OEF is higher in Case B (59.3 %) than in Case A (38.3%).

Statistical analysis

All statistical analyses were performed using the R software program.²⁶ Continuous variables are expressed as mean ± SD; categorical variables are expressed as n (%). A value of p < 0.05 was considered as statistically significant.

For imaging characteristics, brain atrophy was defined as deep (enlargement of the ventricles) or peripheral (enlargement of the gyri) and rated on a subjective scale of 0 to 3 (0 = absent, 1 = mild, 2 = moderate, 3 = severe).²⁷ We divided the patients into two groups based on presence or absence of the WMH (Fazekas grade 0 vs. grade 1) and the pathologic brain atrophy (grade 0 vs. grade 1 and 2).

On the basis of the SLEDAI and BILAG index scores, we categorised the SLE patients into a severe group (SLEDAI score ≥ 11 or BILAG index score ≥ 6) and a non-severe group (SLEDAI score < 11 or BILAG index score < 6) according to the previous reports.^28,29 We also divided the patients with NPS and without NPS into two groups according to their neuro BILAG (nBILAG) index scores (A, B and C vs. D and E).¹⁶ For treatment, patients were divided into two groups based on whether they had a history of pulse steroid therapy. For the analyses, age, duration of SLE, WBC, PLT, Hb, Hct, ESR, CH50, IgG and anti-dsDNA antibody were treated as continuous variables.

We compared the clinical and imaging characteristics, including the QSM-OEF between SLE patients and HS, and between the severe and non-severe groups (as assessed by the SLEDAI and BILAG index scores) using the χ² test and Fisher's exact test for categorical variables and the two-sample t-test or Mann–Whitney U test for continuous variables, as appropriate.

For the QSM-OEF, we calculated the sensitivity, specificity, and area under the curve (AUC) for the prediction of the SLE disease activity (severe group vs. non-severe group). For this analysis, we used cutoff values in a receiver operating characteristic (ROC) curve analysis, which were determined by using the Youden index.

Pearson's or Spearman's correlation coefficients were calculated to examine bivariate correlations. Independent predictors were identified through a multiple linear regression analysis to identify the factors (patient's characteristics, imaging characteristics, SLE-related factors and vascular risk factors) related to QSM-OEF in SLE patients.

Results

As shown in Table 1, there were no statistically significant differences in the age and sex distribution of the patients between SLE patients and HS. QSM-OEF was significantly higher in SLE (51.3 ± 10.1) than that in HS (40.5 ± 3.7) (p < 0.001) (Table 1).

Table 1.

Imaging and clinical characteristics of SLE and healthy subjects.

		SLE (n = 42)	HS (n = 20)	p
Clinical characteristics	Age, ave. ± SD	38.7 ± 15.8	40.7 ± 13.0	0.623
	Sex, female (%)	38 (90)	19 (95)	0.549
	Alcohol, n (%)	2 (5)	3 (15)	0.329
	Smoking, n (%)	3 (7)	1 (5)	0.753
	Obesity, n (%)	1 (2)	0 (0)	0.495
Imaging characteristics	QSM-OEF (%), ave. ± SD	51.3 ± 10.1	40.5 ± 3.7	<0.001
	Atrophy, n (%)	16 (38)	0 (0)	0.001
	WML, n (%)	14 (33)	2 (10)	0.051
Vascular risk factors	HT, n (%)	5 (12)	2 (10)	0.111
	HL, n (%)	3 (7)	0 (0)	0.705
	DM, n (%)	2 (5)	0 (0)	0.329
SLE-related factors	Disease duration, month, ave. ± SD	7.5 ± 7.9	NA	NA
	NP symptom, n (%)	7 (17)	NA	NA
	Disease activity score, n
	SLEDAI severe ≧ 11/non severe < 11	16 (38)/26 (62)	NA	NA
	BILAG index severe ≧ 6/non severe < 6	18 (43)/24 (57)	NA	NA
	History of pulse steroid therapy, n (%)	10 (24)	NA	NA

SLE: systemic lupus erythematosus; HS: healthy subject; SD: standard deviation; QSM: quantitative susceptibility mapping; OEF: oxygen extraction fraction; WML: white matter lesion; HT: hypertension; HL: hyperlipidemia; DM: diabetes mellitus; NP: neuropsychiatric; SLEDAI: Systemic Lupus Erythematosus Disease Activity Index; BILAG: British Isles Lupus Assessment Group; NA: not applicable.

Table 2 shows the results of comparison between the severe and non-severe groups. For the SLEDAI and BILAG index scores, QSM-OEF was significantly higher in the severe group than that in the non-severe group (61.4 ± 9.7 vs. 46.1 ± 5.1, p < 0.001, 59.1 ± 11.2 vs. 46.5 ± 5.4, p < 0.001) (Table 2). The severe group demonstrated significantly lower WBC, PLT, Hb, Hct and CH50 and significantly higher ESR that those in the non-severe group (Table 2).

Table 2.

The clinical and imaging characteristics for the subgroup comparisons in SLE patients.

		SLEDAI < 11	SLEDAI ≧ 11		BILAG index score < 6	BILAG index score ≧ 6
		n = 28	n = 14	p value	n = 26	n = 16	p
Clinical characteristics	Age, ave. ± SD	36.6 ± 14.0	42.9 ± 18.8	0.231	38.1 ± 13.8	39.7 ± 19.1	0.753
	Sex, female (%)	26 (93)	12 (86)	0.469	24 (92)	14 (88)	0.617
	Alcohol, n (%)	1 (4)	1 (7)	0.619	1 (4)	1 (6)	0.73
	Smoking, n (%)	1 (4)	2 (14)	0.213	2 (8)	1 (6)	0.864
	Obesity, n (%)	1 (4)	0 (0)	0.486	1 (4)	0 (0)	0.44
Imaging characteristics	QSM-OEF (%), ave. ± SD	46.1 ± 5.1	61.6 ± 9.7	<0.001	46.5 ± 5.4	59.1 ± 11.2	<0.001
	Atrophy, n (%)	10 (36)	6 (43)	0.663	10 (38)	6 (38)	0.952
	WML, n (%)	10 (36)	4 (29)	0.653	11 (42)	3 (19)	0.121
Vascular risk factors	HT, n (%)	2 (7)	3 (21)	0.186	2 (8)	3 (19)	0.294
	HL, n (%)	3 (11)	0 (0)	0.213	3 (12)	0 (0)	0.166
	DM, n (%)	1 (4)	1 (7)	0.619	1 (4)	1 (6)	0.73
SLE-related factors	Disease duration, month, ave. ± SD	8.0 ± 7.6	6.5 ± 8.4	0.55	8.7 ± 7.6	5.6 ± 8.2	0.226
	NP symptom, n (%)	2 (7)	5 (36)	0.019	2 (8)	5 (31)	0.048
	Disease activity score, ave. ± SD
	SLEDAI	2.2 ± 3.3	16.6 ± 5.0	<0.001	2.2 ± 3.7	14.9 ± 6.3	<0.001
	BILAG index	2.6 ± 3.8	17.9 ± 8.9	<0.001	1.6 ± 1.9	17.6 ± 8.1	<0.001
	History of pulse steroid therapy, n (%)	5 (18)	5 (36)	0.21	5 (19)	5 (31)	0.387
	Blood examinations, ave. ± SD
	WBC (10³/µL)	5.7 ± 2.5	4.1 ± 1.8	0.032	6.0 ± 2.5	3.8 ± 1.2	0.002
	PLT (10⁵/µL)	217.7 ± 58.7	163.6 ± 42.5	0.004	255.5 ± 51.2	157.6 ± 46.6	<0.001
	Hb (g/dL)	12.1 ± 1.5	10.8 ± 1.5	0.009	12.2 ± 1.6	10.9 ± 1.3	0.007
	Hct (%)	36.4 ± 4.5	31.7 ± 5.0	0.003	36.6 ± 4.7	32.0 ± 4.5	0.003
	ESR (mm/h)	26.0 ± 23.5	41.1 ± 26.6	0.066	24.7 ± 23.4	41.3 ± 25.6	0.037
	CH50 (U/ml)	48.4 ± 12.2	35.1 ± 12.0	0.002	48.2 ± 12.7	37.1 ± 12.4	0.009
	IgG (mg/dL)	14.4 ± 4.3	14.4 ± 5.9	0.99	13.7 ± 4.0	15.6 ± 5.9	0.224
	anti-dsDNA antibody (IU/ml)	37.9 ± 85.4	77.8 ± 126.2	0.232	37.5 ± 86.8	73.6 ± 120.4	0.266

SLEDAI: systemic lupus erythematosus disease activity index; BILAG: British isles lupus assessment group; SD: standard deviation; QSM: quantitative susceptibility mapping; OEF: oxygen extraction fraction; WML: white matter lesion; HT: hypertension; HL: hyperlipidemia; DM: diabetes mellitus; NP: neuropsychiatric; WBC: white blood cell; PLT: platelet; Hb: hemoglobin; Hct: hematocrit; ESR: erythrocyte sedimentation rate; CH50: 50% hemolytic unit of complement; IgG: immunoglobin G; anti-dsDNA antibody: anti-double stranded DNA antibody.

The ROC analysis suggested optimal cutoff levels for the QSM-OEF (49.5 for SLEDAI and 51.3 for BILAG score) for prediction of the SLE disease activity. Therefore, the sensitivity, specificity, and AUC were 100.0% (14/14), 75.0% (21 of 28), and 0.946 for the SLEDAI, and 75.0% (12/16), 84.6% (22/26), and 0.849 for the BILAG score, respectively.

As shown in Table 3, QSM-OEF was positively correlated with the SLE disease activity score in the Spearman's correlation analysis (SLEDAI score: ρ = 0.663, p < 0.001 and BILAG index score: ρ = 0.606; p < 0.001) and the presence of NPS in the Pearson's correlation analysis (ρ = 0.340, p = 0.028) (Table 3). In the multiple linear regression analyses for QSM-OEF, SLEDAI and NPS were independently associated with QSM-OEF (For SLEDAI, standardized β = 0.426, p = 0.016 and standardized β = 6.148, p = 0.029, R² = 0.701). BILAG index score was not independently associated with QSM-OEF, but NPS with QSM-OEF (For BILAG index, standardized β = 0.252, p = 0.098 and standardized β = 6.834, p = 0.020, R² = 0.671) (Table 3).

Table 3.

The univariate and multiple linear regression analyses for activity index (SLEDAI, BILAG index score).

		Univariate correlations		Multiple linear regression analyses
				SLEDAI		BILAG index
		ρ	p	Standardized β	p	Standardized β	p
Clinical characteristics	Age	0.011	0.946	–	–	–	–
	Sex	0	1	–	–	–	–
	Alcohol	0.055	0.728	–	–	–	–
	Smoking	−0.006	0.968	–	–	–	–
	Obesity	0.058	0.715	–	–	–	–
Imaging characteristics	Atrophy	0.061	0.703	–	–	–	–
	WML	−0.025	0.875	–	–	–	–
Vascular risk factors	HT	0.203	0.197	–	–	–	–
	HL	−0.24	0.125	–	–	–	–
	DM	0.111	0.485	–	–	–	–
SLE-related factors	Disease duration	−0.049	0.76	–	–	–	–
	NP symptom	0.34	0.028	6.148	0.029	6.834	0.02
	Disease activity score
	SLEDAI	0.663	<0.001	0.426	0.016	–	–
	BILAG index	0.606	<0.001	–	–	0.252	0.098
	History of pulse steroid therapy	0.166	0.293
	Blood examinations
	WBC	−0.308	0.047	−0.137	0.787	−0.014	0.979
	PLT	−0.408	0.007	−0.006	0.758	−0.009	0.648
	Hb	−0.618	<0.001	−0.189	0.936	−0.022	0.993
	Hct	−0.665	<0.001	−1.022	0.152	−1.066	0.157
	ESR	0.299	0.055	−0.109	0.057	−0.094	0.119
	CH50	−0.495	<0.001	−0.13	0.18	−0.175	0.079
	IgG	−0.115	0.467	–	–	–	–
	anti-dsDNA antibody	0.216	0.17	–	–	–	–

SLEDAI: Systemic Lupus Erythematosus Disease Activity Index; BILAG: British Isles Lupus Assessment Group; WML: white matter lesion; HT: hypertension; HL: hyperlipidemia; DM: diabetes mellitus; NP: neuropsychiatric; WBC: white blood cell; PLT: platelet; Hb: hemoglobin; Hct: hematocrit; ESR: erythrocyte sedimentation rate; CH50: 50% hemolytic unit of complement; IgG: immunoglobin G; anti-dsDNA antibody: anti-double stranded DNA antibody.

Discussion

In this study, OEF measurements using QSM of MRI were significantly higher in SLE patients than those in HS. Despite previous descriptions of cerebral blood flow (CBF) and its patterns in SLE,^30–32 the characteristics of cerebral oxygen metabolism have been unclear. Thus, the ability to image brain oxygenation using QSM-OEF offers a non-invasive means to understand brain conditions in the SLE patients, including disruptions to the oxygen supply. Moreover, the novel finding was that OEF was independently and significantly associated with SLEDAI and NPS—which would expect that the SLE-associated systemic inflammatory processes leading to peripheral endothelial dysfunction in SLE brain,^33–36 which are a reported brain pathophysiology in SLE patients.

A physiologic change in Hct content affects QSM-OEF measurements. This effect may be important in pathologic conditions, such as anaemia, in which QSM-OEF is known to increase. In this study, the Hb and Hct level in the patients with a high SLE score was low, approaching the level of anaemia, which would result in elevated OEF. However, it is important to note that, in the multiple linear regression analyses for QSM-OEF, the SLEDAI was independently associated with QSM-OEF, but not Hct and Hb values. These results may indicate that, in the SLE patients, the QSM-OEF map may reflects abnormal oxygen metabolism rather than anaemia.

When an artery is narrowed or occluded to compensate for decreasing cerebral perfusion pressure (CPP), brain arterioles dilate to reduce vascular resistance in attempts to maintain constant CBF (autoregulation).^37,38 As CPP continues to decrease and CBF is maximally increased, a second compensatory mechanism results in an increase in the fraction of total oxygen extracted from blood (OEF)^39,40 to maintain a close to normal oxygen metabolism. This phenomenon termed ‘misery perfusion’⁴¹ represents ischaemic tissue with elevated OEF. For the mechanism for the increased OEF in this study, we spiculate that similar pathogenesis may occur in the brain of SLE patients. This hypothesis is supported by many previous studies, which demonstrated small vessel injury due to vasculitis in SLE brain.⁴² A recent preliminary study using 7T-MRI demonstrated that minute punctate or linear hyperintensity lesions with/without contrast enhancement indicating microvascular inflammation were frequently found in patients with NPSLE.⁴³ Moreover, many previous studies based on ¹⁸FDG PET, H2 ¹⁵⁰PET and SPECT show GM hypometabolism and hypoperfusion in NPSLE patients.^{30–32,44,45} On the other hand, the previous ¹⁸FDG PET study showed the glucose hypermetabolism was specifically elicited by correlation with disease activity in SLE.⁴⁶ Because inflammation increases glucose transporter expression,⁴⁷ increased ¹⁸FDG uptake would be expected in the presence of inflammation. Therefore, as a different mechanism for our results, other than small vessel injury, the increased oxygen demand would be postulated to be attributable to ischemic state in SLE patients. The previous studies using SPECT or MR perfusion-weighted imaging showed increased cerebral blood volume (CBV) in SLE patients.^48–50 The increased CBV reported in stroke may indicate intact vascular beds with potentially salvageable areas in the penumbra of the lesion.⁵¹ Furthermore, increases in CBV with associated decreased arterial resistances and increased vasodilation can be a sign of vasomotor instability, a compensatory mechanism indicating resolving injury or a reduction of compensatory capacity in response to further ischaemia or injury.⁵² The previous study also showed that cerebral arterial resistances and impedances are significantly lower in SLE compared with controls, which were also consistent with increased arteriolar dilation and increased CBF. Thus, in SLE patients, we assume that the condition with increased CBV may cause an increased OEF.

SLE fundamentally involves inflammation, which is difficult to detect using conventional MRI. Although many studies with conventional MRI have supplied ample evidence of brain pathology in SLE patients, such as cerebral infarction, brain atrophy and multifocal GM and/or WM lesions,⁵ these MR findings may indicate that pathological conditions of the brain occur as a result of SLE-related inflammation. Furthermore, previous investigators have considered that NPSLE may cause changes in brain metabolism and blood flow, which can be detected using several radiographic techniques, including PET,⁵³ SPECT,^54,55 and advanced MRI techniques such as perfusion-weighted MRI, fMRI and MR spectroscopy.⁵⁶ However, there are controversial results in the literature concerning brain metabolism and blood flow in NPSLE brain. Although there are widespread reports of hypoperfusion in NPSLE, cerebral hypermetabolism and hyperperfusion in SLE have been noted in a number of previous reports using SPECT, PET or magnetic resonance perfusion-weighted imaging.^48–50 Thus, the etiology of and basis for NPSLE-associated brain lesions remains uncertain, and an imaging biomarker of disease status has been desired.

Our results may show importance of QSM-OEF as an imaging biomarker of disease activity in SLE, although the elevated OEF in SLE patients was pathophysiologically relatively small. To our knowledge, there has only been one study to evaluate an imaging biomarker for the disease activity in SLE.¹² In this study using PET, an association was found between the SLEDAI score and increased ¹⁸FDG uptake in the WM of SLE patients. Because inflammation increases the expression of glucose transporter, the authors interpret the hypermetabolism of the WM as evidence of acute inflammation. However, PET is limited by its rigid technical requirements and high cost.⁵⁷ Thus, our less invasive assessment with QSM-OEF has clinical significance in which it may provide a simple method for investigating disease activity in SLE patients.

Our study has some limitations that should be acknowledged when interpreting the results. First, the main limitation of this study was the lack of cases with contrast-enhanced MR perfusion study. Therefore, the relationship between OEF and perfusion measurement (CBF or CBV) could not be proved. In determining the mechanism for the increased OEF in this study, blood flow measurement was an important factor. Second, we had a relatively small sample size recruited from only one institution which may represent sampling bias. Third, subject selection was also limited by the study's retrospective design. Because many patients received therapy with anti-platelet, anti-hypertensive and anti-lipemic drugs before undergoing MR imaging, we cannot rule out the possibility that these treatments affected the results. Fourth, we used the SLEDAI as a disease activity index in the SLE patients. The SLEDAI was not used as measurement of the degree of CNS involvement alone, but rather as a marker of global disease activity in various organs. Fifth, for the calculation of the QSM-OEF in this study, automated measurement of ROIs was performed at two-dimensional position coordinate (on a slice at the level of centrum semiovale). It would be better to apply standard method such as three-dimensional stereotaxic ROI, which was used in the previous paper regarding the QSM-OEF measurement.⁵⁸ Finally, we excluded patients with brain abnormalities on MRI such as large territory infarction and WMH (Fazekas grade 2 and 3) to simplify investigations. Thus, our study included few patients with severe NPS which could affect the results.

In conclusion, we found the elevated OEF of the SLE brain, which may indicate ‘misery perfusion’ in diffuse brain areas in SLE patients. Our results may offer new insights into the characterisation of hemodynamic impairment of the SLE brain and a better understanding of brain injury evidence and altered brain physiology in SLE. Furthermore, OEF in SLE patients is specifically associated with systemic disease activity. This suggests that the elevated OEF is an indication of reactive changes of the small vessels that may occur as a result of neuroinflammatory activity. Further research should be performed to investigate the longitudinal associations among OEF, disease activity, clinical symptoms and the prognosis of SLE to fully ascertain whether OEF can be used as an imaging biomarker for monitoring the disease activity, an early indicator of SLE, and as a prognostic predictor in SLE patients. Furthermore, predictors of the development of brain lesions on MRI in SLE patients also would be of clinical value.

Footnotes

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors' contributions

MM, SK, KK, YW and YK planned and designed the study and wrote the manuscript. MM and SK analyzed the data. SI and YT collected study participants and clinical date. All authors reviewed the manuscript.

References

O'Connor

Musher

. Central nervous system involvement in systemic lupus erythematosus: a study of 150 cases. Arch Neurol 1966; 14: 157–164.

Sibbitt

Jr Sibbitt

Jr Brooks

, et al. Magnetic resonance imaging and brain histopathology in neuropsychiatric systemic lupus erythematosus. Semin Arthritis Rheum 2010; 40: 32–52.

Kim

Choi

, et al. Primary antiphospholipid antibody syndrome: neuroradiologic findings in 11 patients. Korean J Radiol 2000; 1: 5–10.

Golstein

Meyer

Bourgeois

, et al. Neurological manifestations of systemic lupus erythematosus: role of antiphospholipid antibodies. Clin Exp Rheumatol 1992; 11: 373–379.

Kaichi

Kakeda

Moriya

, et al. Brain MR findings in patients with systemic lupus erythematosus with and without antiphospholipid antibody syndrome. Am J Neuroradiol 2014; 35: 100–105.

Zivadinov

Shucard

Hussein

, et al. Multimodal imaging in systemic lupus erythematosus patients with diffuse neuropsychiatric involvement. Lupus 2013; 22: 675–683.

Gibbs

Leenders

Wise

, et al. Evaluation of cerebral perfusion reserve in patients with carotid-artery occlusion. Lancet 1984; 11: 310–314.

Baron

Bousser

Rey

, et al. Reversal of focal “misery-perfusion syndrome” by extra-intracranial arterial bypass in hemodynamic cerebral ischemia. A case study with 15O positron emission tomography. Stroke 1981; 12: 454–459.

Grubb

Jr Derdeyn

Fritsch

, et al. Importance of hemodynamic factors in the prognosis of symptomatic carotid occlusion. JAMA 1998; 280: 1055–1060.

10.

Yamauchi

Higashi

Kagawa

, et al. Is misery perfusion still a predictor of stroke in symptomatic major cerebral artery disease?. Brain 2012; 135: 2515–2526.

11.

Johnson

Richardson

. The neurological manifestations of systemic lupus erythematosus: a clinical–pathological study of 24 cases and review of the literature. Medicine 1986; 47: 337–369.

12.

Ramage

Fox

Brey

, et al. Neuroimaging evidence of white matter inflammation in newly diagnosed systemic lupus erythematosus. Arthritis Rheum 2011; 63: 3048–3057.

13.

Kudo

Liu

Murakami

, et al. Oxygen extraction fraction measurement using quantitative susceptibility mapping: comparison with positron emission tomography. J Cereb Blood Flow Metab 2016; 36: 1424–1433.

14.

Hochberg

. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1997; 40: 1725.

15.

American College of Rheumatology Ad Hoc Committee on Systemic Lupus Erythematosus Response Criteria. The American College of Rheumatology response criteria for systemic: erythematosus clinical trials. Arthritis Rheum 2004; 50: 3418–3426.

16.

Isenberg

Rahman

Allen

, et al. BILAG 2004. Development and initial validation of an updated version of the British Isles Lupus Assessment Group's disease activity index for patients with systemic lupus erythematosus. Rheumatology 2005; 44: 902–906.

17.

Fazekas

Kleinert

Offenbacher

, et al. The morphologic correlate of incidental punctate white matter hyperintensities on MR images. Am J Neuroradiol 1991; 12: 915–921.

18.

Liu

Surapaneni

Lou

, et al. Cerebral microbleeds: burden assessment by using quantitative susceptibility mapping. Radiology 2012; 262: 269–278.

19.

Liu

de Rochefort

, et al. Morphology enabled dipole inversion for quantitative susceptibility mapping using structural consistency between the magnitude image and the susceptibility map. Neuroimage 2012; 59: 2560–2568.

20.

Liu

Wisnieff

Lou

, et al. Nonlinear formulation of the magnetic field to source relationship for robust quantitative susceptibility mapping. Magn Reson Med 2013; 69: 467–476.

21.

Liu

Khalidov

de Rochefort

, et al. A novel background field removal method for MRI using projection onto dipole fields. NMR Biomed 2011; 24: 1129–1136.

22.

Kressler

de Rochefort

Liu

, et al. Nonlinear regularization for per voxel estimation of magnetic susceptibility distributions from MRI field maps. IEEE Trans Med Imaging 2010; 29: 273–281.

23.

Pauling

Coryell

. The magnetic properties and structure of hemoglobin, oxyhemoglobin and carbonmonoxyhemoglobin. Proc Natl Acad Sci U S A 1936; 22: 210–216.

24.

Kudo

Sasaki

Ogasawara

, et al. Difference in tracer delay–induced effect among deconvolution algorithms in CT perfusion analysis: quantitative evaluation with digital phantoms. Radiology 2009; 251: 241–249.

25.

Kudo

Sasaki

Østergaard

, et al. Susceptibility of T max to tracer delay on perfusion analysis: quantitative evaluation of various deconvolution algorithms using digital phantoms. J Cereb Blood Flow Metab 2011; 31: 908–912.

26.

R Core Team. R: A language and environment for statistical computing: The R foundation for statistical Computing. Available online at http://www.R-project.org/.

27.

Farrell

Chappell

Armitage

, et al. Development and initial testing of normal reference MR images for the brain at ages 65–70 and 75–80 years. Eur Radiol 2009; 19: 177–183.

28.

Mosca

Bombardieri

. Assessing remission in systemic lupus erythematosus. Clin Exp Rheumatol 2006; 24: 99–104.

29.

Kalunian

Kim

Xie

, et al. Impact of standard of care treatments and disease variables on outcomes in systemic lupus erythematosus trials: analysis from the Lupus Foundation of America Collective Data Analysis Initiative. Eur J Rheumatol 2016; 3: 13–19.

30.

Yoshida

Shishido

Kato

, et al. Evaluation of cerebral perfusion in patients with neuropsychiatric systemic lupus erythematosus using 123I–IMP SPECT. Ann Nucl Med 2007; 21: 151–158.

31.

Lin

Wang

Yen

, et al. Technetium–99m–HMPAO brain SPECT in systematic lupus erythematosus with CNS involvement. J Nucl Med 1997; 38: 1112–1115.

32.

Russo

Gilday

Laxer

, et al. Single photon emission computed tomography scanning in childhood systemic lupus erythematosus. J Rheumatol 1998; 25: 576–582.

33.

Rhiannon

. Systemic lupus erythematosus involving the nervous system: presentation, pathogenesis, and management. Clin Rev Allergy Immunol 2008; 34: 356–360.

34.

Roman

Shanker

Davis

, et al. Prevalence and correlates of accelerated atherosclerosis in systemic lupus erythematosus. N Engl J Med 2003; 349: 2399–2406.

35.

Waterloo

Omdal

Sjöholm

, et al. Neuropsychological dysfunction in systemic lupus erythematosus is not associated with changes in cerebral blood flow. J Neurol 2001; 248: 595–602.

36.

Kobayashi

Watanabe

Seino

, et al. Quantitative imaging of cerebral blood flow using SPECT with 123I–iodoamphetamine in neuropsychiatric systemic lupus erythematosus. J Rheumatol 2003; 30: 2075–2076.

37.

MacKenzie

McGeorge

Graham

, et al. Effects of increasing arterial pressure on cerebral blood flow in the baboon: influence of the sympathetic nervous system. Pflugers Arch 1979; 378: 189–195.

38.

Rapela

Green

. Autoregulation of canine cerebral blood flow. Circ Res 1964; 15: 205–212.

39.

Powers

Grubb

Jr Darriet

, et al. Cerebral blood flow and cerebral metabolic rate of oxygen requirements for cerebral function and viability in humans. J Cereb Blood Flow Metab 1985; 5: 600–608.

40.

Powers

. Cerebral hemodynamics in ischemic cerebrovascular disease. Ann Neurol 1991; 29: 231–240.

41.

Baron

Bousser

Comar

, et al. Noninvasive tomographic study of cerebral blood flow and oxygen metabolism in vivo. Eur Neurol 1981; 20: 273–284.

42.

Osborn

. Osborn's brain: Imaging, pathology and anatomy, 1st ed. Philadelphia: Amirsys and Lippincott Williams & Wilkins, 2012.

43.

Murata

Sasaki

, et al. Detection of cerebral microvascular lesions using 7 T MRI in patients with neuropsychiatric systemic lupus erythematosus. Neuroreport 2015; 26: 27–32.

44.

Kao

Lan

ChangLai

, et al. The role of FDG-PET, HMPAO-SPET and MRI in the detection of brain involvement in patients with systemic lupus erythematosus. Eur J Nucl Med 1999; 26: 129–134.

45.

Kovacs

Urowitz

Gladman

, et al. The use of single photon emission computerized tomography in neuropsychiatric SLE: a pilot study. J Rheumatol 1995; 22: 1247–1253.

46.

Ramage

Fox

Brey

, et al. Neuroimaging evidence of white matter inflammation in newly diagnosed systemic lupus erythematosus. Arthritis Rheum 2011; 63: 3048–3057.

47.

Love

Tomas

Tronco

, et al. FDG PET of infection and inflammation. Radiographics 2005; 25: 1357–1368.

48.

Gasparovic

Roldan

Sibbitt

Jr , et al. Elevated cerebral blood flow and volume in systemic lupus measured by dynamic susceptibility contrast magnetic resonance imaging. J Rheumatol 2010; 37: 1834–1843.

49.

Lee

Joo

Bang

, et al. Basal ganglia hyperperfusion in a patient with systemic lupus erythematosus–related parkinsonism. Neurology 2004; 63: 395–396.

50.

Khubchandani

Viswanathan

Desai

. Unusual neurologic manifestations (I): Parkinsonism in juvenile SLE. Lupus 2007; 16: 572–575.

51.

Hatazawa

Shimosegawa

Toyoshima

, et al. Cerebral blood volume in acute brain infarction. Stroke 1999; 30: 800–806.

52.

Smith

. Perioperative uses of transcranial perfusion monitoring. Neurosurg Clin N Am 2008; 19: 489–502.

53.

Lee

Park

Lee

, et al. The efficacy of brain 18 F-fluorodeoxyglucose positron emission tomography in neuropsychiatric lupus patients with normal brain magnetic resonance imaging findings. Lupus 2012; 21: 1531–1537.

54.

Castellino

Padovan

Bortoluzzi

, et al. Single photon emission computed tomography and magnetic resonance imaging evaluation in SLE patients with and without neuropsychiatric involvement. Rheumatology 2008; 47: 319–323.

55.

Zhang

Zhu

Zhang

, et al. Diagnostic value of single–photon–emission computed tomography in severe central nervous system involvement of systemic lupus erythematosus: a case–control study. Arthritis Rheum 2005; 53: 845–849.

56.

Popescu

Kao

AHA

. Neuropsychiatric systemic lupus erythematosus. Curr Neuropharmacol 2011; 9: 449–457.

57.

Meyer

Schober

Stoppe

, et al. Cerebral involvement in systemic lupus erythematosus (SLE): comparison of positron emission tomography (PET) with other imaging methods. Psychiatry Res 1989; 29: 367–368.

58.

Uwano

Kudo

Sato

, et al. Noninvasive assessment of oxygen extraction fraction in chronic ischemia using quantitative susceptibility mapping at 7 Tesla. Stroke 2017; 48: 2136–2141.

Evaluation of oxygen extraction fraction in systemic lupus erythematosus patients using quantitative susceptibility mapping

Abstract

Keywords

Introduction

Material and methods

Approval

Subjects

MRI

Data processing of OEF

Statistical analysis

Results

Discussion

Footnotes

Funding

Declaration of conflicting interests

Authors' contributions

References