The disease

A stroke is a “sudden” circulatory disruption in the brain which causes the death of brain tissue and thus the failure of bodily functions (e.g. paralysis). Tissue death can be caused by a lack of blood flow, e.g., by clogging of a blood vessel, or by bleeding in the brain. Strokes are the most common cause of acquired disability in adulthood, and one third of stroke patients die within one year (1).

So far there is only one effective therapy - the dissolution of blood clots using rt-PA (recombinant tissue-specific plasminogen activator), but only 80-85% of strokes, the so-called ischemic strokes, are caused by a vascular occlusion. This means that the other patients, namely those with a bleeding-related stroke, cannot be treated with it. In addition, rt-PA must be administered within 4.5 hours of the occurrence of the stroke to be effective. Due to these limitations, treatment with rt-PA can only be used in about 5% of patients (2).

The search for new therapies, in particular for so-called neuroprotectives (drugs that protect nerve tissue), has so far led to more than 4,000 publications in which the neuroprotective properties of more than 700 drugs could apparently be proven in animal experiments (3). Other studies refer to up to 1,000 treatment methods tested on animals, but they did not have the desired effect on humans (4). Since the approval of rt-PA for the treatment of stroke in 1996, there has been no significant progress, although stroke has been investigated in so-called animal models for over 150 years (5).

So-called animal models

Experiments performed on rats are by far the most frequent, but mice, cats, dogs, pigs, monkeys (6, 7) and fish (7) are also used. Rats are used for research primarily because rat husbandry is simple and cheap, and rats can easily be manipulated genetically.

Since the 1970s, more and more new "animal models" have been developed in which a stroke is artificially triggered by the occlusion of an artery supplying the brain. The mortality rate can be up to 40%; individual studies even report mortality rates as high as 85% (6).

Many "animal models" require the skull to be opened via a so-called craniotomy. After the craniotomy, one of the arteries supplying the brain (usually the middle cerebral artery) is exposed and blood clots or small particles are flushed in (thromboembolic model) to close or narrow the blood vessel. Other methods used to close the blood vessels include electric current (electrocoagulation model) or injection of endothelin-1, a hormone that causes blood vessels to constrict.

The electrocoagulation method can cause jaw problems in rats, necessitating a switch to soft food. This in turn leads to uncontrolled growth of the teeth, which then have to be ground down regularly (8) - an additional physical and psychological burden on the animals.

The endothelin-1 model is associated with particularly severe suffering for the animals, because a vinyl cannula (a small plastic tube) is inserted through a hole drilled in the skull. This cannula is sewn in and left there for a few days. Then endothelin-1 is injected via the tube and the animals experience the stroke while being fully conscious (9, 10). The mortality rate is as high as 15% (8).

But even the models in which the skull is not opened represent a considerable burden for the animals, and result in pain. In the so-called intraluminal thread model, a surgical thread is introduced via the cervical artery into the middle cerebral artery (MCA) until it is blocked by the thread. Depending on the specific research question, the thread is left there for 30 minutes to several hours. As a result, a large area of tissue in the brain is no longer supplied with blood and oxygen. It is not uncommon for the animals to suffer cerebral hemorrhage and overheating, or to die as a result of the extensive stroke (11, 12).

In the photochemical model, the skull is irradiated after the injection of a light-sensitive dye. As a result, the platelets clump together and form a clot that blocks the vessel. In other models, the artery is closed with clips, hooks, ligatures, cuffs, or by injecting thrombin (a clotting factor) (8).

Although these models have been continuously developed further, even animal scientists admit that no single "animal model" can fully mimic the situation of a stroke patient (6, 11, 12, 13, 14).

Drug testing

The stroke models are not only used to investigate the histological, physiological, and biochemical mechanisms of the disease, but also to test new neuroprotective (brain-protecting) drugs. However, so far none of the drugs that have been deemed successful in animal experiments have been effective in humans. Many drugs prove ineffective in clinical trials with patients. Often clinical studies had to be discontinued because drugs that had been classified as safe in animal experiments resulted in undesired, sometimes dangerous adverse effects, or even worsened the patient’s outcome (e.g. the degree of disability after the stroke).

Clinical trials with the new drugs Selfotel and Aptiganel had to be stopped because mortality among the patients treated with the drug was higher than in the placebo group (the group of patients who received a dummy drug) (15, 16). Patients who took Selfotel also suffered from hallucinations, confusion, and impaired consciousness, including coma (15). Another study with albumin found that patients treated with this substance had a 2.4-time greater risk of having a cerebral hemorrhage; the most common side effect was pulmonary edema (accumulation of fluid in the lungs) (17).

Reasons for the failure of stroke research

Poor study quality:

The overall quality of animal studies is poor (4, 18, 19, 20, 21). Important quality features for medical studies are: the random allocation of patients to the different groups in order to prevent biased selection and thus measurement errors (so-called randomization), as well as blinding, i.e. patients and/or the investigators do not know which drug they are receiving (this prevents the individual groups from being evaluated differently depending on the expectations or hopes of the investigators). However, only 36% of animal studies are randomized and only 29% are blinded (22). For this reason, drugs and other therapeutic approaches are incorrectly classified as effective, or side effects are incorrectly assessed.

For example, more than 100 studies on cold treatment (hypothermia) of stroke patients have been published and more than 3,000 animals have been "consumed" for the corresponding research (23). In order to keep brain damage to a minimum in the event of a stroke, the patient's body temperature should be cooled to less than 35°C. Although animal experiments showed positive effects of hypothermia, this could not be confirmed in clinical studies (24).

Discrepancies between “animal models” and patients:

The animals used in the laboratory are generally young and have no pre-existing conditions or risk factors, while stroke patients are typically between 60 and 70 years old and suffer from one or more underlying health conditions (vascular calcification, high blood pressure, diabetes, obesity, elevated blood lipid levels, etc.). Stroke patients are therefore much more complex than animal models can ever be (25). Gender, age, and genetic factors also play important roles in stroke (and in other conditions too, of course).

Although there are now "animal models" in which individual risk factors have been introduced (e.g., spontaneously hypertensive rats, ApoE-deficient mice) (8), these animals, in contrast to stroke patients, are genetically manipulated or artificially made ill (e.g. the kidney arteries are clamped to induce high blood pressure). This is not comparable to the complex course of the human disease (5), in which vascular calcification, caused by diabetes or high blood pressure, develops over years, sometimes decades. Many patients also take medications that affect their risk for a stroke, the chance of recovery, and response to stroke treatment.

There are also numerous anatomical differences. The human brain, for example, has a much larger proportion of white matter (the portion of the brain in which the extensions of the neurons, the so-called axons, are found) than the rodent brain (10% vs. 50% (26, 11)), which is particularly important when testing drugs that specifically target the white or gray matter (the part of the brain that contains the neuronal cell bodies).

The blood supply to the cerebral arterial circle, which supplies blood to the brain, is mainly provided via the carotid artery in animals, but additionally via the vertebral artery (A. vertebralis) in humans (27).

Even if these differences appear minimal, in highly complex organisms such as humans and animals, even tiny, seemingly negligible differences can have dramatic effects (25). Such differences are important, for example, when it comes to bypass circulations (i.e., vessels that take over the supply of a brain region when another artery is blocked) or in surgical access routes and treatment methods. In addition to differences between humans and animals (see above), there are also differences within an animal species. Various studies have even been able to identify differences between individual rat and mouse breeds (28, 29).

Different read-outs:

To assess the success of a therapy in animal experiments, usually the infarction area, i.e., the area with the damaged brain tissue, is measured. A reduction in the size of the infarct after administration of the drug is considered a success. In clinical studies on patients, on the other hand, various functional values are measured, in particular with the help of rating scales, which can be used to assess the limitations affecting the patient since the stroke and see how the patients are coping with them in everyday life (23, 26).

The reduction of the infarction area does not necessarily correspond with an improvement in the neurological deficits (e.g., paralysis, disturbances of sensation, speech and memory disorders) (26). Nor does a major infarction necessarily result in severe neurological deficits, which can be more dependent on the localization of the infarction (for example, even a mini-infarction in the brainstem can have a fatal outcome).

Many important neurological functions cannot be tested in animals, such as speech or memory disorders that many stroke patients suffer from (5), or alterations in personality. In order to compensate for this deficit, some researchers apply various behavioral tests. A well-known example is the so-called Morris water maze: the animal (usually a rat) is placed in a container filled with turbid water, with markers attached to the walls. There is a platform, not visible under the water surface, which the rat is supposed to find using the landmarks. Among other things, this test is supposed to examine memory (26). The grid walking test (the animals have to balance on horizontal ladders and the examiners measure the number of missteps) or the wire or grid hanging test are used to check coordination and mobility. For example, a mouse is hung by its paws on a grid or wire and the examiner records how long the animal can hold itself there before falling down. (30).

Without exception, all of these tests are dependent on the compliance (i.e. the "cooperation") of the so-called test animals, which can be doubted when one considers that the lives of these animals are usually characterized by pain, isolation, boredom, and fear, and the fact that researchers often purposely starve the animals before some of these tests to “encourage” cooperation. Nobody can judge whether a rat takes longer to find the platform in the Morris water maze because of stroke-induced limitations, or because it is scared, in pain, starving, or simply not in the mood.

Different time frames and dosages:

In order to avoid toxic effects, patients in clinical trials are usually given much lower doses of the test drug, although these vary greatly between individual studies. A dosage that was effective in animals might not be appropriate for patients (26). Humans might  tolerate much higher or much lower doses than rats or other animals.

In animal experiments, the drugs are often administered before (!) or during the stroke. The realism of such studies should be questioned, as these time points are not feasible in everyday clinical practice and would require physicians to have clairvoyant abilities.

Although a stroke patient needs to obtain treatment as soon as possible, this cannot be done until the diagnosis has been confirmed. This usually requires imaging to differentiate between a stroke caused by a lack of oxygen and a stroke caused by bleeding (because the therapies are different), or to rule out other causes of the symptoms. This takes time.

While in animal experiments the animals are administered the test drug an average of ten minutes after inducing the stroke (ranging from 60 minutes before (!) to 360 minutes after) (31), in clinical studies it takes an average of 12 hours before the patients are included in a study and the treatment is started (32). In the case of hypothermia treatment, cooling was started after an average of 0 minutes (!) in the animal experiment, i.e., immediately after inducing the stroke. The best effectiveness of hypothermia was found when the treatment was started before or during vascular occlusion. Such timelines cannot be applied in everyday clinical practice, since sometime has to be allowed for transport, diagnostics, and initial treatment. Moreover, cooling takes about 4 hours in a human but only about 20 minutes in a rat (23).

Conclusion

Reasons for the unsuccessful attempts to transfer results from animal experiments to patients include the sometimes alarmingly poor quality of animal studies, the significant and insurmountable differences between patients and "experimental" animals in terms of anatomy, physiology, and previous illnesses, inconsistent and different assessments of the results, as well as different time frames and dosages in animal experiments and clinical studies.

15 May 2018
Stephanie Gräwert, medical practitioner

References

(1) https://www.kompetenznetz-schlaganfall.de/176.0.html, Abruf am 15.5.2018
(2) Canazza A, Minati L et al.: Experimental models of brain ischemia: a review of techniques, magnetic resonance imaging, and investigational cell-based therapies. Frontiers in Neurology, 2014, 5: 1-14
(3) Macleod MR, O´Collins T et al.: Pooling of Animal Experimental Data Reveals Influence of Study Design and Publication Bias. Stroke, 2004, 35:1203-1208
(4) Greek R, Menache A: Systematic Reviews of Animal Models: Methodology versus Epistemology. Int J Med Sci, 2013, 10(3): 206-221
(5) Wiebers DO, Adams HP, Whisnant JP: Animal Models of Stroke: Are They Relevant to Human Disease? Stroke, 1990, 21(1): 1-3
(6) Howells DW, Porritt MJ et al.: Different strokes for different folks: the rich diversity of animal models of focal cerebral ischemia. Journal of Cerebral Blood Flow & Metabolism, 2010, 30: 1412-1431
(7) Walcott BP, Peterson RT: Zebrafish models of cerebrovascular disease. Journal of Cerebral Blood Flow & Metabolism, 2014, 34: 571-577
(8) Macrae IM: Preclinical stroke research – advantages and disadvantages of the most common rodent models of focal ischemia. British Journal of Pharmacology, 2011, 164: 1062-1078
(9) Macrae IM, Robinson MJ et al.: Endothelin-1-Induced Reductions in Cerebral Blood Flow: Dose Dependency, Time Course, and Neuropathological Consequences, Journal of Cerebral Blood Flow & Metabolism, 1993, 13: 276-284
(10) Sharke J, Ritchie IM, Kelly PAT: Perivascular Microapplication of Endothelin-1: A New Model of Focal Cerebral Ischaemia in the Rat. Journal of Cerebral Blood Flow & Metabolism, 1993, 13: 865-871
(11) Willing AE: Experimental Models. Help or Hindrance. Stroke, 2009, 40 (suppl 1): 152-154
(12) Morgenthaler P, Meisel A: Do stroke models model stroke? Disease Models & Mechanisms, 2012, 5: 718-725
(13) Bacigaluppi M, Comi G, Hermann DM: Animal Models of Ischemic Stroke. Part One: Modeling Risk Factors. The Open Neurology Journal, 2010, 4: 26-33
(14) Bacigaluppi M, Comi G, Hermann DM: Animal Models of Ischemic Stroke. Part Two: Modeling Cerebral Ischemia, 2010, 4: 34-38
(15) Davis SM, Lees KR et al.: Selfotel in Acute Ischemic Stroke. Possible Neurotoxic Effects of an NMDA Antagonist. Stroke, 2000, 31: 347-354
(16) Albers GW, Goldstein LB, Hall D, Lesko LM: Aptiganel hydrochloride in acute ischemic stroke: a randomized controlled trial. JAMA (2001); 286(21): 2673-82 (Abstract)
(17) Ginsberg MD, Palesch YY et al.: High-dose albumin treatment for acute ischaemic stroke (ALIAS): a phase 3, randomized, double-blind, placebo-controlled trial. Lancet Neurol, 2013, 12(11): 1049-1058
(18) Philip M, Benatar M et al.: Methodological Quality of Animal Studies of Neuroprotective Agents Currently in Phase II/III Acute Ischemic Stroke Trials. Stroke, 2009, 40: 577-581
(19) Kilkenny C, Parsons N et al.: Survey of the Quality of Experimental Design, Statistical Analysis and Reporting of Research Using Animals. PlosOne, 2009, 4(11):e7824
(20) Pedder H, Vesterinen HM et al.: Systematic Review and Meta-Analysis of Interventions Tested in Animal Models of Lacunar Stroke. Stroke, 2014, 45: 563-570
(21) Agoston DV: Bench-to-bedside and bedside back to bench; seeking a better understanding of the acute pathophysiological process in severe traumatic brain injury. Frontiers in Neurology, 2015, 6(47): 1-6
(22) Fisher M, Feuerstein G et al.: Update of the Stroke Therapy Academic Industry Roundtable Preclinical Recommendations. Stroke, 2009, 40(6): 2244-2250
(23) van der Worp HB, Sena ES et al.: Hypothermia in animal models of acute ischemic stroke: a systematic review and meta-analysis. Brain, 2007, 130: 3063-3074
(24) Piironen K, Tiainen M et al.: Mild Hypothermia After Intravenous Thrombolysis in Patients With Acute Stroke. Stroke, 2014, 45: 486-491
(25) Joffe AR, Bara M et al.: Expectations for methodology and translation of animal research: a survey of health care workers. BMC Medical Ethics, 2015, 16: 29
(26) Cheng YD, Al-Khoury L, Zivin JA: Neuroprotection for Ischemic Stroke: Two Decades of Success and Failure. The Journal of the American Society for Experimental NeuroTherapeutics, 2004, 1(1): 36-45
(27) Ashwini CA, Shubhia R, Jayanthi KS: Comparative anatomy of the circle of Willis in man, cow, sheep, goat, and pig. Neuroanatomy, 2008, 7: 54-65
(28) Barone FC, Knudsen DJ et al.: Mouse Strain Differences in Susceptibility to Cerebral Ischemia Are Related to Cerebral Vascular Anatomy. Journal of Cerebral Blood Flow & Metabolism, 1993, 13: 683-692
(29) Bardutzky J, Shen Q et al.: Differences in Ischemic Lesion Evolution in Different Rat Strains Using Diffusion and Perfusion Imaging. Stroke, 2005, 36(9): 2000-2005
(30) Balkaya M, Kröber JM et al.: Assessing post-stroke behavior in mouse models of focal ischemia. Journal of Cerebral Blood Flow & Metabolism, 2013, 33: 330-338
(31) van der Worp HB, de Haan P et al.: Methodological quality of animal studies on neuroprotection in focal cerebral ischaemia. J Neurol, 2005, 252(9): 1108-14
(32) Kidwell CS, Liebeskind DS et al.: Trends in Acute Ischemic Stroke Trials Through the 20th Century. Stroke, 2001, 32: 1349-1359