Using MRI to Observe Increased Venous Flow Collateralization in Subjects with Anomalous Jugular Veins

SEAN KUMAR SETHI¹, Giacomo Gadda², Ana M. Daugherty³, David T. Utriainen¹, Jing Jiang¹, Naftali Raz³, and Ewart Mark Haacke⁴

¹The MRI Institute of Biomedical Research, Detroit, MI, United States, ²Physics, University of Ferrara, Ferrara, Italy, ³Department of Gerontology, Wayne State University, Detroit, MI, United States, ⁴Biomedical Engineering, Wayne State University, Detroit, MI, United States

Synopsis

We have established in previous works that a subset of multiple sclerosis (MS) patients show abnormal structure and flow in the internal jugular veins (IJV) when measured with MRI. In this retrospective analysis, we classified and compared extracranial venous collateral flow in MS and normal control samples using MR venography and Phase-contrast flow quantification with a large, standardized dataset. Over 50% of the MS cohort shows a jugular anomaly. The stenotic-MS group shows reduced Type I venous flow compared to healthy controls and non-stenotic MS, while having elevated Type II and Type III flows.

Purpose

We have established in previous works that a subset of multiple sclerosis (MS) patients show abnormal structure and flow in the internal jugular veins (IJV) when measured with magnetic resonance imaging (MRI). In this retrospective analysis, we classified and compared extracranial venous collateral flow in MS and normal control samples using MRI with a large, standardized dataset¹. We hypothesize that MS patients with anomalous IJVs would have elevated collateral venous flow compared to controls.

Methods

A group of 559 MS subjects and 106 healthy control (HC) subjects were imaged with 3T MRI scanners from four different imaging sites. Phase-contrast flow quantification (PCFQ) and MRV imaging were collected to image blood flow and anatomy of the extracranial vessels². With MR venography, IJVs were classified into stenotic (ST) and non-stenotic (NST) groups based on absolute cross-sectional area threshold^2,3,4,5. Individual and total vessel flow, as well as venous flow normalized to total arterial flow (tA) were quantified using in-house software with a custom phase-unwrapping algorithm. Veins were classified by cranial drainage into three collateral types: I: IJVs, II: paraspinal veins, and III: superficial veins. Differences in stenosis prevalence in HC and MS were assessed by chi-square test with significance of p<0.05. Statistical comparisons of flow properties were done in MS, HC, and between group subcategories using a General Linear Model. Flow indices including bilateral normalized Type I, II, and III venous flow at both the C2/C3 and C5/C6 neck levels were used as dependent variables. First, possible differences between cervical level was tested in a 2 (cervical level) x 3 (vessel type) x 2 (MS vs Control). Additional analyses per cervical level, treating vessel type as a 3-level repeated measure were conducted as warranted. Omnibus effects and complex interactions were further decomposed in post hoc via paired- or independent-sample t-tests as appropriate, and Pearson correlations. These post-hoc analyses were bootstrapped (5000 draws, 100% of the observed sample) to estimated bias-corrected 95% confidence intervals (CI). Additionally, ROC curve analyses were done for discriminating among HC, NST-MS, ST-MS groups with respect to Type I, II, and III venous flow. Prior to data processing, all data processors met a ICC2 statistic of >0.9 for flow/cardiac cycle measurements of IJV, cerebral arteries, as well as Total Type I, II and III venous flows.

Results

In the MS group, 346/559 (61%) classified as ST while 12/60 (20%) HC classified as ST (X2=59.3, p<0.05). Forty-six cases did not include venography in their protocol. The differential flow between vessels differed between cervical levels (level x type, F(2,660)=17.4, p<0.001), the group differences by vessel type were different between the two levels (level x type x group, F(2,660)=26.1, p<0.001). Therefore, all additional analyses were done separately by cervical level. In a secondary GLM treating vessel type as a 3-level repeated measure, group differences between stenotic MS, non-stenotic MS and HC were assessed including age and sex as covariates. Bivariate interactions were tested and found to be not significant (all p>0.21) and were removed from the model. For the C5/C6 level, the net flow between jugular (type I), paraspinal (Type 2) and superficial flow (Type 3) was significantly different. Flow in Type I vessels was greater than in Type II and III and Type II vessels had the least amount of flow. Further, similar results were observed at the C2/C3 level. P-values <0.001 for both levels. For C5/C6, NST-MS and HC had similar flow in all vessel types. In comparison to either of these groups, stenotic MS had less flow in type I vessels, and greater flow in type II and type III vessels. This pattern was also observed at the C2/C3 neck level with all p-values <0.001. See Table 1 for a comparison of mean flow values between groups by vessel type. Age (p=0.80) and sex (p=0.70) were unrelated to differences in total net flow.

Conclusions

We have shown that over 50% of the MS cohort shows a jugular anomaly, while only 20% of the HC show a jugular anomaly. But most critically, we were able to show the similarities between the non-stenotic MS group and the healthy controls in terms of flow properties for both cervical levels, and all three collateral flow types. The stenotic group shows markedly reduced Type I venous flow compared to HC and NST, while having elevated Type II and Type III venous flow.

Acknowledgements

The authors would like to thank MR Innovations India, Robert Loman, MD, and Imran Saqib for assistance with data processing. Dr. Phil Levy from Wayne State University for usage of healthy control data; and the Annette Funicello Research Fund and the Center for Neurological Diseases for assistance with funding. This work was supported in part by a grant from the National Institute on Aging, R37-AG011230 to NR.

References

1. Pacurar EE, Sethi SK, Habib C, et al. Database integration of protocol-specific neurological imaging datasets. Neuroimage 2015

2. Haacke EM, Feng W, Utriainen D, et al. Patients with multiple sclerosis with structural venous abnormalities on MR imaging exhibit an abnormal flow distribution of the internal jugular veins. J Vasc Interv Radiol. 2012;23(1):60-8 e1-3.

3. Feng W, Utriainen D, Trifan G, et al. Characteristics of flow through the internal jugular veins at cervical C2/C3 and C5/C6 levels for multiple sclerosis patients using MR phase contrast imaging. Neurological research. 2012;34(8):802-9.

4. Feng W, Utriainen D, Trifan G, Sethi S, Hubbard D, Haacke EM. Quantitative flow measurements in the internal jugular veins of multiple sclerosis patients using magnetic resonance imaging. Rev Recent Clin Trials. 2012;7(2):117-26.

5. Sethi SK, Utriainen DT, Daugherty AM, et al. Jugular Venous Flow Abnormalities in Multiple Sclerosis Patients Compared to Normal Controls. J Neuroimaging. 2014.

Figures

Table 1: Means and standard deviations are reported; bias-corrected bootstrapped 95% confidence intervals of the mean are shown in parentheses. 95% confidence intervals that do not overlap support a group difference at p < 0.05.

Figure 1A: Plot showing total normalized venous flow at C5/C6 vs. C2/C3 for Type I vessels. ROC optimum values which discriminate between ST-MS and HC are plotted as dotted lines on each graph.

Table 2: ROC analysis results showing optimal normalized venous flow thresholds between ST-MS and HC, as well as the number of cases that fall below the thresholds for each group and subgroup (MS, NST-MS, ST-MS, HC, and ST-HC).

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)

2710