Do pre-clinical multiple sclerosis models allow us to measure neurodegeneration and clinical progression?

KEYWORDS:

• Brain atrophy
• clinical trials
• cuprizone
• experimental autoimmune encephalomyelitis
• neurodegeneration
• progressive multiple sclerosis

1. General aspects

Besides demyelination, damage to neurons drives disease progression in multiple sclerosis (MS) patients. Different disease courses exist. Relapsing remitting MS (RRMS) affects about 85% of all MS patients. RRMS means that symptoms appear (i.e.. a relapse) and then fade away either partially or completely between new attacks (i.e.. remitting). Secondary progressive MS (SPMS) is characterized by chronically progressive clinical worsening overtime. This progressive disease course usually follows RRMS. In about 15% of patients, the disease is characterized by worsening neurologic function from the onset of symptoms, without early relapses or remissions, called primary progressive MS (PPMS). While inflammatory demyelination of the white matter is the pathological correlate of acute relapses, irreversible damage to neurons is the pathological correlate of disease progression. Of note, virtually every element of the neuron can be destroyed including cell soma, dendritic spines, synapses and neurites [1].

2. MS treatment options

It is widely accepted that activation of the peripheral immune system, which results in inflammatory demyelination, causes acute relapses. Consequently, therapeutic strategies include modalities aiming at antagonizing the various immune elements that are involved in this multifaceted cascade. While corticosteroids are used to reduce the acute inflammation that spikes during a relapse, disease-modifying drugs are used to reduce the relapse frequency. Since first described by Charcot almost two centuries ago, our knowledge about the pathology and pathophysiology of MS has progressed considerably. Not only do we know better about the pathophysiology of inflammatory lesion development but also the different risk factors, clinical subtypes, and how to attenuate the pathophysiological processes leading to inflammatory damage of the CNS. The success cannot be better seen than in the field of RRMS with more than six medications got approved by the US FDA and EMA since 2010. However, the primary and secondary progressive disease courses (i.e. SPMS and PPMS) remain a considerable challenge. In fact, at the time of this writing, just one medication (Novantrone®; mitoxantrone) has been approved by the FDA for SPMS [2]. For PPMS, ocrelizumab (Ocrevus®) is the only FDA-approved disease-modifying therapy [3]. Currently, there is no cure for MS. In this editorial we take a critical look at animal models used to develop new therapeutic modalities for progressive MS.

3. How appropriately do we use preclinical MS models?

Although animal models are artificial, and findings using these models have to be interpreted with some caution, they are infinitely valuable to study mechanisms of inflammatory lesion development. White matter inflammatory demyelination blocks axonal conduction, and relief of inflammation results in recovery. Experimental autoimmune encephalomyelitis (EAE) is the most commonly used tool to study this important aspect. In EAE, experimental animals (commonly rodents) are immunized with a myelin-related peptide administered in a strong adjuvant, usually complete Freund’s adjuvant. The combination of the peptide and the immunized mouse strain determines the disease course, idem est relapsing or chronic disease. Another model which recently has become increasingly popular to study mechanisms and consequences of demyelination is the cuprizone model, in which primary oligodendrocyte apoptosis leads to innate immune activation and finally demyelination. A third model which can be used to study progressive MS is the Theiler’s murine encephalomyelitis virus (TMEV) model [4]. During the chronic disease phase, the virus infects glial cells and macrophages, which leads to inflammatory demyelination with oligodendrocyte apoptosis and axonal degeneration. In these models, neurodegeneration occurs to a variable extent [5]. The main difference between these models is that in EAE and TMEV, tissue damage is mainly triggered by the acquired immune system, whereas in the cuprizone model innate immune activation results in tissue damage.

To test for efficacy of any therapeutic intervention, one needs appropriate outcome measurements. One important point shall be highlighted here: preclinical studies apparently use other outcome measurements compared to those used in SPMS and PPMS clinical trials. In clinical trials, the two main primary outcome measurements are (i) progression of clinical disability and (ii) cerebral atrophy, measured by MRI [6,7]. The traditionally used primary clinical outcome measure is the Expanded Disability Status Scale (EDSS). Although the EDSS is still the mainstay for assessing treatment efficacy, nowadays several studies use much more sophisticated approaches such as the MS Functional Composite (MSFC). The MSFC comprises quantitative functional measures of three key clinical dimensions of MS: leg function/ambulation, arm/hand function, and cognitive function. Delay in time to sustained accumulated disability, and improvement of change in brain volume both indicate treatment success [8]. Both measurements, clinical disability and brain volume, are believed to depend on the extent of neurodegeneration.

What are the behavioral/clinical outcome measurements used in preclinical studies? In the cuprizone model, behavioral/clinical deficit measurements are not frequently applied, although some studies have shown that cuprizone-intoxication induces such deficits [9]. In EAE, the most commonly used outcome measurement is the extent of motor deficits, which is mainly driven by spinal cord pathology. Each mouse is graded daily and assigned a score ranging from 0 to 5. Parameters include limp tail or hind limb weakness when EAE is mild, and partial or complete hind limb paralysis in severe EAE cases. Of note, this measurement is neither very sensitive nor objective. Cognitive or sensory deficits are usually not measured during preclinical studies.

What about neurodegeneration? Is this aspect frequently evaluated in preclinical studies? To come straight to the point: The answer is no! Common readouts in the cuprizone model are extent of oligodendrocyte destruction, demyelination, and gliosis, whereas extent of inflammatory demyelination in the spinal cord, measured by (immuno-)histochemistry, is a common readout in EAE studies. Brain atrophy, which is an important measurement in clinical studies, is almost never evaluated in preclinical trials, although MRI-based protocols have been published for both, EAE [10] and the cuprizone model [11]. Equally, loss of neurons, spines, and dendrites is almost never determined in preclinical trials. To conclude, outcome measurements in clinical and preclinical trials are fundamentally different. But why is that the case?

4. What can we do better?

First, it is a matter of time. It is important to remember that it takes time for the spinal cord and brain to shrink. Most preclinical studies only follow mice for some weeks or even less. Preclinical studies should be carried out for longer periods in order to see effects of the inflammatory damage on neuron loss and atrophy. Another limitation of currently applied EAE protocols is the severity of the model. In EAE, mice become completely paralyzed and neuronal damage becomes so rapid and extensive that recovery cannot or hardly occur. The window of intervention is, therefore, so narrow that it is difficult to intervene with neuroprotective agents in EAE. This highlights another important point in this article – that there is a need for preclinical models that are less severe. In the ideal case, mice should show deficits, but are still able to ambulate. This way, mice can be maintained in good health for a longer periods of time, which allows for natural aging of the disease, creating a larger time window for therapeutic intervention. Milder preclinical models would be more amenable to measures of complex gait analysis [12], and would allow us to study cognitive deficits.

Second, it is a matter of limitations in the applied histological protocols. In ‘conventional’ (immuno-) histological studies, cell profiles on a cut section are analyzed. However, after slicing, the cell profile can misrepresent the true properties of the cell. A bigger object (say, a pyramidal neuron) has more chance of appearing on a tissue section than a smaller object (say, an interneuron). More precisely, if a given treatment induces neuronal swelling, the probability of each neuron to appear on a tissue section is greater for swollen versus non-swollen neurons. In consequence, cell swelling might lead to the false-positive result: increase in cell number. The opposite is true if cell shrinkage occurs in response to treatment. Of note, protocols to measure reliably neuronal cell loss as well as brain atrophy are readily available namely design-based stereology [13,14]. Other methods to quantify the extent of neurodegeneration include silver stain impregnation for the visualization of dendrites and spines, or the preparation of epon embedded, toluidine blue stained sections for the visualization of myelin and axons. Another elegant and equally powerful method is recording of axon conduction via electrophysiology [15].

Another issue which should be kept in mind is that most imaging and pathological MS studies are performed in the forebrain. By contrast, most EAE studies focus on spinal cord lesions. Equally this limitation can be compassed by applying appropriate EAE protocols [16,17].

5. Conclusion

To conclude, the currently applied measurements during, especially, EAE studies were effective for the development of anti-inflammatory drugs. For successfully developing progressive MS therapies, we have to focus on complex clinical/functional outcome measurements combined with in depth characterization of neurodegeneration. Furthermore, there is an urgent need for milder preclinical models where mice are followed for longer times. In that event only will we be able in the future to effectively screen for novel compounds which show efficacy in SPMS and PPMS clinical trials.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.