Why brain atrophy measurements matter

Mar 2, 2017

In Multiple Sclerosis (MS), irreversible neurological impairment is being caused by inflammation and neurodegenerative processes. Characteristics are lesions in the central nervous system (area of acute injury caused by inflammation) and abnormal brain atrophy, or brain volume loss (BVL). As the disease is incurable, treatment of MS is focused on disease-modifying therapy (DMT). For each individual patient, effectiveness of therapy is monitored closely and therapeutic decisions are guided by critical parameters, such as relapse rate, disability progression, and lesion development.

Recent studies have prompted debates about the clinical relevance of BVL over time as a measure to quantify neuro-degeneration, and subsequently the use of current brain atrophy as guidance in therapeutic decisions 1, 2, 3 What is clear from this debate is the need for a uniform and validated algorithm that could incorporate brain volume changes in therapeutic decisions alongside clinical data and MRI lesion changes4.

NEDA-3 to NEDA-4?

With the emergence of new and more effective disease-modifying therapies (DMT), the management of multiple sclerosis became increasingly complex the latest years. Several data challenged the old treatment paradigm of simply reducing relapse rates and the consequences of relapses in MS patients. Consequently, the treatment of Relapse-Remitting MS (RRMS) was redefined with the target “No Evidence of Disease Activity” (NEDA). Until recently, therapy decisions of RRMS-patients were based on three (NEDA-3) related measures of disease activity:

• - No relapses

• - No disability progression

• - No MRI activity (new or enlarging T2 lesions or Gd-enhancing lesions) 5, 6

  
  

Newer studies and data suggest incorporation of another prognostic parameter in the NEDA guidelines 4, 6,7. NEDA-4 also includes change in brain volume over time for treatment decisions as BVL is accelerated in (all stages of) MS and correlates with disability and cognitive decline. Thorough investigation of clinical trials further revealed that there is an association between treatment effects on BVL and on disability13. Including BVL in therapy decisions implies appropriate MRI monitoring strategies, analysis and algorithms, etc. which currently is a point of some debate...

Why you should be including MRI brain atrophy measurements in MS clinical practice

Including BVL as an MS biomarker is not generally accepted by specialists yet. This was emphasized recently with some publications in Multiple Sclerosis Journal1,3...in which proponents and opponents gave their argumentation. Barkhof1 underscores the lack of validated cut-off values for BV measures of cerebral atrophy. He addresses several variability inducing parameters in assessing BV (including within-patient biological variability – age, hydration, hormones, etc. – small technical variations – movement artifacts, etc. and technical variability – scanner type, methods of analysis, etc. -) to conclude that brain atrophy measurements are not yet sufficiently reliable to guide treatment in individual MS patients. “And even with a robust technical implementation, there would still be a need for a clear algorithm,” Barkhof states. Chard (2) tends to agree with Barkhof and highlights the mismatch between the established role of brain atrophy measures in early-phase treatment trials, and the possible role they could play in tailored treatments for individual MS patients.

On the other hand, Zivadinov and colleagues3 clearly plead that the assessment of brain atrophy has to become accepted by regulatory agencies: “Nowadays, the interpretation of brain atrophy results from MS clinical trials has important implications for individual treatment choice selection both in academic and in community-based MS centers, hence there is an imminent need for incorporating brain atrophy assessments at the individual patient level for treatment monitoring.” The authors do acknowledge the challenges in brain atrophy assessment at the individual patient level, and they strongly agree that we have to continue working on solving these challenges. But, as studies show that BVL provides an important additional value in determining or explaining the effect of DTM, they believe that BV assessment has to be incorporated into treatment monitoring today.

This was also shown in another recent publication7, where a group of MS neurologists and neuroradiologists concluded that adding the monitoring of BVL to NEDA provides a better treatment and monitoring strategy.

How does icobrain ms contribute to the adoption of BVL monitoring in clinical practice?

icometrix developed icobrain ms to respond to the growing evidence and the need for inclusion of brain atrophy into the routine clinical practice to follow individual patients with MS. icobrain ms calculates brain and lesion volumes and uses MRI scans of multiple time points to detect volume changes. In different (independent) studies 8,9,10, it has been shown that icobrain ms serves as a valid medical software tool to robustly assess brain atrophy. At icometrix, we are convinced that brain atrophy is a critical and valuable parameter for treatment of MS in different stages.

We believe NEDA-4 is a strong concept in MS. However, as also raised by Barkhof, there are two things to consider when including brain atrophy in routine MS clinical practice:

- Biological variability

- Technical variability

With reliable algorithms and validated analysis software these issues can be overcome. First of all, it has been shown that gray matter is less likely affected by pseudo-atrophy and other non-demyelinating causes (dehydration, smoking, diurnal variation) than white matter (WM)11,12, hence most attention has been focused on either whole-brain or grey matter volume quantification. Next, the measurement error of the method needs to be sufficiently low, in order to draw meaningful conclusions for individual patients. Several independent studies with icobrain ms software 8-10 provide promising results indicating icobrain ms serves as a software solution that is sufficiently robust to uniformly assess brain volume. Although further validation is still ongoing, we are confident that icobrain ms can already contribute significantly to disease monitoring.

Low measurement error with MRI biomarkers

icobrain ms is built upon years of intense research and development work together with physicians to reduce the measurement error of the method.

• only structures that can be measured sufficiently reliable are reported (only whole brain and grey matter atrophy / no hippocampal atrophy)

• longitudinal techniques are used to directly measure the tissue change (atrophy)

• lesion filling is incorporated to avoid inconsistent quantification of atrophy

Guidance and standardization of MRI protocols

1. Expert quality control provides guiding for correct interpretation of the measurements.

2. feedback regarding drifting scanner parameters reduces scanner variability

3. feedback on artifacts like motion reduces the small technical variations

Mining the data from previous MRI scans

icobrain ms reports take into account all available previous MRI scans the trend averages out the variability: to illustrate: if two atrophy measurements are performed: the variability is reduced by a factor 2, if three consecutive measurements are performed the variability is reduced by a factor 3

Easy adoption into clinical workflows

The highly dedicated research team of icometrix continuously challenges and improves icobrain ms in order to meet the rising standards in clinical neuroimaging.

References

1. Barkhof (2016) Multiple Sclerosis Journal 1–2 DOI: 10.1177/1352458516649452

2. Chard: (2016) Multiple Sclerosis Journal 1–2 DOI: 10.1177/1352458516656061

3. Zivadinov, Dwyer and Bergsland (2016) Multiple Sclerosis Journal, 1–3 DOI: 10.1177/1352458516649253

4. Kappos L, De Stefano N, Freedman MS, et al. (2016) Mult Scler. Sep;22(10):1297- 305

5. Giovannoni, Turner, Gnanapavan et al. (2011) Multiple Sclerosis and Related Disorders, 4/4; p 329-333

6. NEDA-4 May Predict MS Outcomes Better Than NEDA-3 Neurology Reviews. 2015 December;23(12):35

7. Neurology Reviews. 2015 December;23(12):35(7) Alroughani et al. BMC Neurology (2016) 16:240

8. Wang C, Beadnall H N, Hatton S, et al. (2015) J Neurol Neurosurg Psychiatry

9. Smeets D, Ribbens A, Sima DM et al. (2016) Brain Behav. Jul 19;6(9)

10. Lysandropoulos AP, Absil J, Metens T, et al. (2016) Brain Behav. Jan 12;6(2)

11. Arevalo O et al. (2019) J Comput Assist Tomogr. Jan/Feb;43(1);1-12

12. Dieleman N et al. (2017) Neuroimage Clin. Sep 6;16-507-513

13. Maria Pia Sormani (2014) Ann Neurol. Jan;75(1):43-9.